[rocRoller][hipblaslt]Use separate array for flags in rocRoller StreamK by liu-yiqian · Pull Request #3177 · ROCm/rocm-libraries

liu-yiqian · 2025-12-05T20:09:55Z

Motivation

We need to use a separate scratch space for flags in StreamK and guarantees they are all 0s before and after the kernel launch.

Technical Details

KernelGraph changes in AddStreamK graph transform to reset the flags to 0 after the DoWhile loop.
Use a allocator per policy system to replace the single allocator for scratch space.
Update the GEMMTest and client to allocate multiple scratch space.

Test Plan

Verify each ScratchPolicy has its own allocator.
Verify the load, store, and reset flags coordinates and control graph are expected after AddStreamK.

Test Result

All unit tests, GEMM tests, client tests should pass.

Submission Checklist

Look over the contributing guidelines at https://github.com/ROCm/ROCm/blob/develop/CONTRIBUTING.md#pull-requests.

math-ci · 2025-12-05T20:14:09Z

Generated Documentation

Link to view Generated Docs.

math-ci · 2025-12-05T20:53:10Z

CodeQL report

Results Summary

Full table of results

Tool	Severity	Code	Location	Line

Links

HTML
Sarif (for download and usage in conjunction with SARIF viewers)

math-ci · 2025-12-05T20:53:52Z

Performance Report for gfx12

Results

Details

@@            Significant (p-val <0.05) Performance Diffs            @@
====================================================================================================
-   0.10% | p=2.5347e-02 
	| 3. FloatsGEMM(M: 7680, N: 8448, K: 8192, alpha: 2, beta: 0.5, types: {'type_A': 'fp8', 'type_B': 'bf8', 'type_C': 'float', 'type_D': 'float', 'type_acc': 'float', 'trans_A': 'N', 'trans_B': 'T', 'scale_A': 'None', 'scaleType_A': 'None', 'scale_B': 'None', 'scaleType_B': 'None', 'scaleBlockSize': -1, 'scaleSkipPermlane': False, 'scaleShuffleTileA': [], 'scaleShuffleTileB': []}, scaleValue_A: 1, scaleValue_B: 1, initMode_A: DataInitMode(Bounded), initMode_B: DataInitMode(Bounded), initMode_C: DataInitMode(Bounded), workgroupMappingDim: -1, workgroupMappingValue: -1)| GEMM(mac_m: 64, mac_n: 64, mac_k: 64, wave_m: 16, wave_n: 16, wave_k: 16, wave_b: 1, workgroup_size_x: 64, workgroup_size_y: 2, workgroupRemapXCC: False, workgroupRemapXCCValue: -1, unroll_x: 0, unroll_y: 0, load_A: BufferToLDSViaVGPR, load_B: BufferToLDSViaVGPR, storeLDS_D: True, betaInFma: True, scheduler: Priority, schedulerCost: LinearWeighted, prefetch: False, prefetchInFlight: 2, prefetchLDSFactor: 0, prefetchMixMemOps: False, loadScale_A: BufferToVGPR, loadScale_B: BufferToVGPR, swizzleScale: False, swizzleTileSize: {'m': 0, 'k': 0, 'n': 0, 'l': 0}, prefetchScale: False, streamK: False, numWGs: 0, streamKTwoTile: False, streamKTwoTileDPFirst: False, architecture: {'ArchString': 'gfx1201', 'Xnack': False, 'Sramecc': False}, matchMemoryAccess: True, version: ) | GEMMProblem(M=7680, N=8448, K=8192, alpha=2, beta=0.5, types={'type_A': 'fp8', 'type_B': 'bf8', 'type_C': 'float', 'type_D': 'float', 'type_acc': 'float', 'trans_A': 'N', 'trans_B': 'T', 'scale_A': 'None', 'scaleType_A': 'None', 'scale_B': 'None', 'scaleType_B': 'None', 'scaleBlockSize': -1, 'scaleSkipPermlane': False, 'scaleShuffleTileA': [], 'scaleShuffleTileB': []}, scaleValue_A=1, scaleValue_B=1, initMode_A='DataInitMode(Bounded)', initMode_B='DataInitMode(Bounded)', initMode_C='DataInitMode(Bounded)', workgroupMappingDim=-1, workgroupMappingValue=-1)GEMMSolution(mac_m=64, mac_n=64, mac_k=64, wave_m=16, wave_n=16, wave_k=16, wave_b=1, workgroup_size_x=64, workgroup_size_y=2, workgroupRemapXCC=False, workgroupRemapXCCValue=-1, unroll_x=0, unroll_y=0, load_A='BufferToLDSViaVGPR', load_B='BufferToLDSViaVGPR', storeLDS_D=True, betaInFma=True, scheduler='Priority', schedulerCost='LinearWeighted', prefetch=False, prefetchInFlight=2, prefetchLDSFactor=0, prefetchMixMemOps=False, loadScale_A='BufferToVGPR', loadScale_B='BufferToVGPR', swizzleScale=False, swizzleTileSize={'m': 0, 'k': 0, 'n': 0, 'l': 0}, prefetchScale=False, streamK=False, numWGs=0, streamKTwoTile=False, streamKTwoTileDPFirst=False, architecture={'ArchString': 'gfx1201', 'Xnack': False, 'Sramecc': False}, matchMemoryAccess=True, version='')
-   0.09% | p=1.7451e-03 
	| 3. FloatsGEMM(M: 7680, N: 8448, K: 8192, alpha: 2, beta: 0.5, types: {'type_A': 'fp8', 'type_B': 'fp8', 'type_C': 'float', 'type_D': 'float', 'type_acc': 'float', 'trans_A': 'N', 'trans_B': 'N', 'scale_A': 'None', 'scaleType_A': 'None', 'scale_B': 'None', 'scaleType_B': 'None', 'scaleBlockSize': -1, 'scaleSkipPermlane': False, 'scaleShuffleTileA': [], 'scaleShuffleTileB': []}, scaleValue_A: 1, scaleValue_B: 1, initMode_A: DataInitMode(Bounded), initMode_B: DataInitMode(Bounded), initMode_C: DataInitMode(Bounded), workgroupMappingDim: -1, workgroupMappingValue: -1)| GEMM(mac_m: 64, mac_n: 64, mac_k: 64, wave_m: 16, wave_n: 16, wave_k: 16, wave_b: 1, workgroup_size_x: 64, workgroup_size_y: 2, workgroupRemapXCC: False, workgroupRemapXCCValue: -1, unroll_x: 0, unroll_y: 0, load_A: BufferToLDSViaVGPR, load_B: BufferToLDSViaVGPR, storeLDS_D: True, betaInFma: True, scheduler: Cooperative, schedulerCost: LinearWeighted, prefetch: False, prefetchInFlight: 2, prefetchLDSFactor: 0, prefetchMixMemOps: False, loadScale_A: BufferToVGPR, loadScale_B: BufferToVGPR, swizzleScale: False, swizzleTileSize: {'m': 0, 'k': 0, 'n': 0, 'l': 0}, prefetchScale: False, streamK: False, numWGs: 0, streamKTwoTile: False, streamKTwoTileDPFirst: False, architecture: {'ArchString': 'gfx1201', 'Xnack': False, 'Sramecc': False}, matchMemoryAccess: True, version: ) | GEMMProblem(M=7680, N=8448, K=8192, alpha=2, beta=0.5, types={'type_A': 'fp8', 'type_B': 'fp8', 'type_C': 'float', 'type_D': 'float', 'type_acc': 'float', 'trans_A': 'N', 'trans_B': 'N', 'scale_A': 'None', 'scaleType_A': 'None', 'scale_B': 'None', 'scaleType_B': 'None', 'scaleBlockSize': -1, 'scaleSkipPermlane': False, 'scaleShuffleTileA': [], 'scaleShuffleTileB': []}, scaleValue_A=1, scaleValue_B=1, initMode_A='DataInitMode(Bounded)', initMode_B='DataInitMode(Bounded)', initMode_C='DataInitMode(Bounded)', workgroupMappingDim=-1, workgroupMappingValue=-1)GEMMSolution(mac_m=64, mac_n=64, mac_k=64, wave_m=16, wave_n=16, wave_k=16, wave_b=1, workgroup_size_x=64, workgroup_size_y=2, workgroupRemapXCC=False, workgroupRemapXCCValue=-1, unroll_x=0, unroll_y=0, load_A='BufferToLDSViaVGPR', load_B='BufferToLDSViaVGPR', storeLDS_D=True, betaInFma=True, scheduler='Cooperative', schedulerCost='LinearWeighted', prefetch=False, prefetchInFlight=2, prefetchLDSFactor=0, prefetchMixMemOps=False, loadScale_A='BufferToVGPR', loadScale_B='BufferToVGPR', swizzleScale=False, swizzleTileSize={'m': 0, 'k': 0, 'n': 0, 'l': 0}, prefetchScale=False, streamK=False, numWGs=0, streamKTwoTile=False, streamKTwoTileDPFirst=False, architecture={'ArchString': 'gfx1201', 'Xnack': False, 'Sramecc': False}, matchMemoryAccess=True, version='')
-   0.20% | p=2.5347e-02 
	| 3. FloatsGEMM(M: 7680, N: 8448, K: 8192, alpha: 2, beta: 0.5, types: {'type_A': 'fp8', 'type_B': 'fp8', 'type_C': 'float', 'type_D': 'float', 'type_acc': 'float', 'trans_A': 'T', 'trans_B': 'T', 'scale_A': 'None', 'scaleType_A': 'None', 'scale_B': 'None', 'scaleType_B': 'None', 'scaleBlockSize': -1, 'scaleSkipPermlane': False, 'scaleShuffleTileA': [], 'scaleShuffleTileB': []}, scaleValue_A: 1, scaleValue_B: 1, initMode_A: DataInitMode(Bounded), initMode_B: DataInitMode(Bounded), initMode_C: DataInitMode(Bounded), workgroupMappingDim: -1, workgroupMappingValue: -1)| GEMM(mac_m: 64, mac_n: 64, mac_k: 64, wave_m: 16, wave_n: 16, wave_k: 16, wave_b: 1, workgroup_size_x: 64, workgroup_size_y: 2, workgroupRemapXCC: False, workgroupRemapXCCValue: -1, unroll_x: 0, unroll_y: 0, load_A: BufferToLDSViaVGPR, load_B: BufferToLDSViaVGPR, storeLDS_D: True, betaInFma: True, scheduler: Sequential, schedulerCost: LinearWeighted, prefetch: False, prefetchInFlight: 2, prefetchLDSFactor: 0, prefetchMixMemOps: False, loadScale_A: BufferToVGPR, loadScale_B: BufferToVGPR, swizzleScale: False, swizzleTileSize: {'m': 0, 'k': 0, 'n': 0, 'l': 0}, prefetchScale: False, streamK: False, numWGs: 0, streamKTwoTile: False, streamKTwoTileDPFirst: False, architecture: {'ArchString': 'gfx1201', 'Xnack': False, 'Sramecc': False}, matchMemoryAccess: True, version: ) | GEMMProblem(M=7680, N=8448, K=8192, alpha=2, beta=0.5, types={'type_A': 'fp8', 'type_B': 'fp8', 'type_C': 'float', 'type_D': 'float', 'type_acc': 'float', 'trans_A': 'T', 'trans_B': 'T', 'scale_A': 'None', 'scaleType_A': 'None', 'scale_B': 'None', 'scaleType_B': 'None', 'scaleBlockSize': -1, 'scaleSkipPermlane': False, 'scaleShuffleTileA': [], 'scaleShuffleTileB': []}, scaleValue_A=1, scaleValue_B=1, initMode_A='DataInitMode(Bounded)', initMode_B='DataInitMode(Bounded)', initMode_C='DataInitMode(Bounded)', workgroupMappingDim=-1, workgroupMappingValue=-1)GEMMSolution(mac_m=64, mac_n=64, mac_k=64, wave_m=16, wave_n=16, wave_k=16, wave_b=1, workgroup_size_x=64, workgroup_size_y=2, workgroupRemapXCC=False, workgroupRemapXCCValue=-1, unroll_x=0, unroll_y=0, load_A='BufferToLDSViaVGPR', load_B='BufferToLDSViaVGPR', storeLDS_D=True, betaInFma=True, scheduler='Sequential', schedulerCost='LinearWeighted', prefetch=False, prefetchInFlight=2, prefetchLDSFactor=0, prefetchMixMemOps=False, loadScale_A='BufferToVGPR', loadScale_B='BufferToVGPR', swizzleScale=False, swizzleTileSize={'m': 0, 'k': 0, 'n': 0, 'l': 0}, prefetchScale=False, streamK=False, numWGs=0, streamKTwoTile=False, streamKTwoTileDPFirst=False, architecture={'ArchString': 'gfx1201', 'Xnack': False, 'Sramecc': False}, matchMemoryAccess=True, version='')
-   0.37% | p=2.5347e-02 
	| 3. FloatsGEMM(M: 7680, N: 8448, K: 8192, alpha: 2, beta: 0.5, types: {'type_A': 'half', 'type_B': 'half', 'type_C': 'half', 'type_D': 'half', 'type_acc': 'half', 'trans_A': 'N', 'trans_B': 'N', 'scale_A': 'None', 'scaleType_A': 'None', 'scale_B': 'None', 'scaleType_B': 'None', 'scaleBlockSize': -1, 'scaleSkipPermlane': False, 'scaleShuffleTileA': [], 'scaleShuffleTileB': []}, scaleValue_A: 1, scaleValue_B: 1, initMode_A: DataInitMode(Bounded), initMode_B: DataInitMode(Bounded), initMode_C: DataInitMode(Bounded), workgroupMappingDim: -1, workgroupMappingValue: -1)| GEMM(mac_m: 64, mac_n: 64, mac_k: 64, wave_m: 16, wave_n: 16, wave_k: 16, wave_b: 1, workgroup_size_x: 64, workgroup_size_y: 2, workgroupRemapXCC: False, workgroupRemapXCCValue: -1, unroll_x: 0, unroll_y: 0, load_A: BufferToLDSViaVGPR, load_B: BufferToLDSViaVGPR, storeLDS_D: True, betaInFma: True, scheduler: Priority, schedulerCost: LinearWeighted, prefetch: False, prefetchInFlight: 2, prefetchLDSFactor: 0, prefetchMixMemOps: False, loadScale_A: BufferToVGPR, loadScale_B: BufferToVGPR, swizzleScale: False, swizzleTileSize: {'m': 0, 'k': 0, 'n': 0, 'l': 0}, prefetchScale: False, streamK: False, numWGs: 0, streamKTwoTile: False, streamKTwoTileDPFirst: False, architecture: {'ArchString': 'gfx1201', 'Xnack': False, 'Sramecc': False}, matchMemoryAccess: True, version: ) | GEMMProblem(M=7680, N=8448, K=8192, alpha=2, beta=0.5, types={'type_A': 'half', 'type_B': 'half', 'type_C': 'half', 'type_D': 'half', 'type_acc': 'half', 'trans_A': 'N', 'trans_B': 'N', 'scale_A': 'None', 'scaleType_A': 'None', 'scale_B': 'None', 'scaleType_B': 'None', 'scaleBlockSize': -1, 'scaleSkipPermlane': False, 'scaleShuffleTileA': [], 'scaleShuffleTileB': []}, scaleValue_A=1, scaleValue_B=1, initMode_A='DataInitMode(Bounded)', initMode_B='DataInitMode(Bounded)', initMode_C='DataInitMode(Bounded)', workgroupMappingDim=-1, workgroupMappingValue=-1)GEMMSolution(mac_m=64, mac_n=64, mac_k=64, wave_m=16, wave_n=16, wave_k=16, wave_b=1, workgroup_size_x=64, workgroup_size_y=2, workgroupRemapXCC=False, workgroupRemapXCCValue=-1, unroll_x=0, unroll_y=0, load_A='BufferToLDSViaVGPR', load_B='BufferToLDSViaVGPR', storeLDS_D=True, betaInFma=True, scheduler='Priority', schedulerCost='LinearWeighted', prefetch=False, prefetchInFlight=2, prefetchLDSFactor=0, prefetchMixMemOps=False, loadScale_A='BufferToVGPR', loadScale_B='BufferToVGPR', swizzleScale=False, swizzleTileSize={'m': 0, 'k': 0, 'n': 0, 'l': 0}, prefetchScale=False, streamK=False, numWGs=0, streamKTwoTile=False, streamKTwoTileDPFirst=False, architecture={'ArchString': 'gfx1201', 'Xnack': False, 'Sramecc': False}, matchMemoryAccess=True, version='')
-   2.06% | p=5.6994e-05 
	| CodeGen(instCount: 40000, instructions: comments)| CodeGen() | CodeGen(instCount: 40000, instructions: comments)
+   1.74% | p=5.6994e-05 
	| CodeGen(instCount: 40000, instructions: simple_mi)| CodeGen() | CodeGen(instCount: 40000, instructions: simple_mi)

Links

math-ci · 2025-12-05T20:53:54Z

Resource Report for gfx12

Results

Details

✔️ No Resource Usage Changes ✔️

Links

math-ci · 2025-12-05T21:43:10Z

Code Coverage Report for gfx942

Summary

Type	Total	Missed	Master Missed	Missed Change	Coverage	Master Coverage	Coverage Change
Lines	60725	11974	11929	45	80.28%	80.30%	-.02%
Functions	5722	1152	1152	0	79.87%	79.86%	.01%
Regions	38824	11115	11001	114	71.37%	71.56%	-.19%
Branches	20931	6130	6088	42	70.71%	70.84%	-.13%

This PR adds/edits 2 newly uncovered lines.

Artifacts

Commit Hashes

math-ci · 2025-12-06T03:09:07Z

Performance Report for gfx942

Results

Details

@@            Significant (p-val <0.05) Performance Diffs            @@
====================================================================================================
+   0.04% | p=2.5347e-02 
	| 3. FloatsGEMM(M: 3072, N: 4096, K: 4096, alpha: 2, beta: 0.5, types: {'type_A': 'float', 'type_B': 'float', 'type_C': 'float', 'type_D': 'float', 'type_acc': 'float', 'trans_A': 'N', 'trans_B': 'N', 'scale_A': 'None', 'scaleType_A': 'None', 'scale_B': 'None', 'scaleType_B': 'None', 'scaleBlockSize': -1, 'scaleSkipPermlane': False, 'scaleShuffleTileA': [], 'scaleShuffleTileB': []}, scaleValue_A: 1, scaleValue_B: 1, initMode_A: DataInitMode(Bounded), initMode_B: DataInitMode(Bounded), initMode_C: DataInitMode(Bounded), workgroupMappingDim: -1, workgroupMappingValue: -1)| GEMM(mac_m: 64, mac_n: 64, mac_k: 64, wave_m: 32, wave_n: 32, wave_k: 2, wave_b: 1, workgroup_size_x: 128, workgroup_size_y: 2, workgroupRemapXCC: False, workgroupRemapXCCValue: -1, unroll_x: 0, unroll_y: 0, load_A: BufferToLDSViaVGPR, load_B: BufferToLDSViaVGPR, storeLDS_D: True, betaInFma: True, scheduler: Priority, schedulerCost: LinearWeighted, prefetch: True, prefetchInFlight: 2, prefetchLDSFactor: 0, prefetchMixMemOps: False, loadScale_A: BufferToVGPR, loadScale_B: BufferToVGPR, swizzleScale: False, swizzleTileSize: {'m': 0, 'k': 0, 'n': 0, 'l': 0}, prefetchScale: False, streamK: False, numWGs: 0, streamKTwoTile: False, streamKTwoTileDPFirst: False, architecture: {'ArchString': 'gfx942', 'Xnack': False, 'Sramecc': True}, matchMemoryAccess: True, version: ) | GEMMProblem(M=3072, N=4096, K=4096, alpha=2, beta=0.5, types={'type_A': 'float', 'type_B': 'float', 'type_C': 'float', 'type_D': 'float', 'type_acc': 'float', 'trans_A': 'N', 'trans_B': 'N', 'scale_A': 'None', 'scaleType_A': 'None', 'scale_B': 'None', 'scaleType_B': 'None', 'scaleBlockSize': -1, 'scaleSkipPermlane': False, 'scaleShuffleTileA': [], 'scaleShuffleTileB': []}, scaleValue_A=1, scaleValue_B=1, initMode_A='DataInitMode(Bounded)', initMode_B='DataInitMode(Bounded)', initMode_C='DataInitMode(Bounded)', workgroupMappingDim=-1, workgroupMappingValue=-1)GEMMSolution(mac_m=64, mac_n=64, mac_k=64, wave_m=32, wave_n=32, wave_k=2, wave_b=1, workgroup_size_x=128, workgroup_size_y=2, workgroupRemapXCC=False, workgroupRemapXCCValue=-1, unroll_x=0, unroll_y=0, load_A='BufferToLDSViaVGPR', load_B='BufferToLDSViaVGPR', storeLDS_D=True, betaInFma=True, scheduler='Priority', schedulerCost='LinearWeighted', prefetch=True, prefetchInFlight=2, prefetchLDSFactor=0, prefetchMixMemOps=False, loadScale_A='BufferToVGPR', loadScale_B='BufferToVGPR', swizzleScale=False, swizzleTileSize={'m': 0, 'k': 0, 'n': 0, 'l': 0}, prefetchScale=False, streamK=False, numWGs=0, streamKTwoTile=False, streamKTwoTileDPFirst=False, architecture={'ArchString': 'gfx942', 'Xnack': False, 'Sramecc': True}, matchMemoryAccess=True, version='')
+   0.41% | p=2.5347e-02 
	| 3. FloatsGEMM(M: 7680, N: 8448, K: 8448, alpha: 2, beta: 0.5, types: {'type_A': 'half', 'type_B': 'half', 'type_C': 'half', 'type_D': 'half', 'type_acc': 'float', 'trans_A': 'N', 'trans_B': 'T', 'scale_A': 'None', 'scaleType_A': 'None', 'scale_B': 'None', 'scaleType_B': 'None', 'scaleBlockSize': -1, 'scaleSkipPermlane': False, 'scaleShuffleTileA': [], 'scaleShuffleTileB': []}, scaleValue_A: 1, scaleValue_B: 1, initMode_A: DataInitMode(Bounded), initMode_B: DataInitMode(Bounded), initMode_C: DataInitMode(Bounded), workgroupMappingDim: -1, workgroupMappingValue: -1)| GEMM(mac_m: 128, mac_n: 128, mac_k: 16, wave_m: 32, wave_n: 32, wave_k: 8, wave_b: 1, workgroup_size_x: 128, workgroup_size_y: 2, workgroupRemapXCC: False, workgroupRemapXCCValue: -1, unroll_x: 0, unroll_y: 0, load_A: BufferToLDSViaVGPR, load_B: BufferToLDSViaVGPR, storeLDS_D: True, betaInFma: True, scheduler: Priority, schedulerCost: LinearWeighted, prefetch: False, prefetchInFlight: 0, prefetchLDSFactor: 0, prefetchMixMemOps: False, loadScale_A: BufferToVGPR, loadScale_B: BufferToVGPR, swizzleScale: False, swizzleTileSize: {'m': 0, 'k': 0, 'n': 0, 'l': 0}, prefetchScale: False, streamK: True, numWGs: 304, streamKTwoTile: False, streamKTwoTileDPFirst: False, architecture: {'ArchString': 'gfx942', 'Xnack': False, 'Sramecc': True}, matchMemoryAccess: True, version: ) | GEMMProblem(M=7680, N=8448, K=8448, alpha=2, beta=0.5, types={'type_A': 'half', 'type_B': 'half', 'type_C': 'half', 'type_D': 'half', 'type_acc': 'float', 'trans_A': 'N', 'trans_B': 'T', 'scale_A': 'None', 'scaleType_A': 'None', 'scale_B': 'None', 'scaleType_B': 'None', 'scaleBlockSize': -1, 'scaleSkipPermlane': False, 'scaleShuffleTileA': [], 'scaleShuffleTileB': []}, scaleValue_A=1, scaleValue_B=1, initMode_A='DataInitMode(Bounded)', initMode_B='DataInitMode(Bounded)', initMode_C='DataInitMode(Bounded)', workgroupMappingDim=-1, workgroupMappingValue=-1)GEMMSolution(mac_m=128, mac_n=128, mac_k=16, wave_m=32, wave_n=32, wave_k=8, wave_b=1, workgroup_size_x=128, workgroup_size_y=2, workgroupRemapXCC=False, workgroupRemapXCCValue=-1, unroll_x=0, unroll_y=0, load_A='BufferToLDSViaVGPR', load_B='BufferToLDSViaVGPR', storeLDS_D=True, betaInFma=True, scheduler='Priority', schedulerCost='LinearWeighted', prefetch=False, prefetchInFlight=0, prefetchLDSFactor=0, prefetchMixMemOps=False, loadScale_A='BufferToVGPR', loadScale_B='BufferToVGPR', swizzleScale=False, swizzleTileSize={'m': 0, 'k': 0, 'n': 0, 'l': 0}, prefetchScale=False, streamK=True, numWGs=304, streamKTwoTile=False, streamKTwoTileDPFirst=False, architecture={'ArchString': 'gfx942', 'Xnack': False, 'Sramecc': True}, matchMemoryAccess=True, version='')
-   2.45% | p=5.6994e-05 
	| 3. FloatsGEMM(M: 7680, N: 8448, K: 8448, alpha: 2, beta: 0.5, types: {'type_A': 'half', 'type_B': 'half', 'type_C': 'half', 'type_D': 'half', 'type_acc': 'float', 'trans_A': 'N', 'trans_B': 'T', 'scale_A': 'None', 'scaleType_A': 'None', 'scale_B': 'None', 'scaleType_B': 'None', 'scaleBlockSize': -1, 'scaleSkipPermlane': False, 'scaleShuffleTileA': [], 'scaleShuffleTileB': []}, scaleValue_A: 1, scaleValue_B: 1, initMode_A: DataInitMode(Bounded), initMode_B: DataInitMode(Bounded), initMode_C: DataInitMode(Bounded), workgroupMappingDim: -1, workgroupMappingValue: -1)| GEMM(mac_m: 128, mac_n: 64, mac_k: 16, wave_m: 32, wave_n: 32, wave_k: 8, wave_b: 1, workgroup_size_x: 128, workgroup_size_y: 2, workgroupRemapXCC: False, workgroupRemapXCCValue: -1, unroll_x: 0, unroll_y: 0, load_A: BufferToLDSViaVGPR, load_B: BufferToLDSViaVGPR, storeLDS_D: True, betaInFma: True, scheduler: Priority, schedulerCost: LinearWeighted, prefetch: False, prefetchInFlight: 0, prefetchLDSFactor: 0, prefetchMixMemOps: False, loadScale_A: BufferToVGPR, loadScale_B: BufferToVGPR, swizzleScale: False, swizzleTileSize: {'m': 0, 'k': 0, 'n': 0, 'l': 0}, prefetchScale: False, streamK: True, numWGs: 304, streamKTwoTile: False, streamKTwoTileDPFirst: True, architecture: {'ArchString': 'gfx942', 'Xnack': False, 'Sramecc': True}, matchMemoryAccess: True, version: ) | GEMMProblem(M=7680, N=8448, K=8448, alpha=2, beta=0.5, types={'type_A': 'half', 'type_B': 'half', 'type_C': 'half', 'type_D': 'half', 'type_acc': 'float', 'trans_A': 'N', 'trans_B': 'T', 'scale_A': 'None', 'scaleType_A': 'None', 'scale_B': 'None', 'scaleType_B': 'None', 'scaleBlockSize': -1, 'scaleSkipPermlane': False, 'scaleShuffleTileA': [], 'scaleShuffleTileB': []}, scaleValue_A=1, scaleValue_B=1, initMode_A='DataInitMode(Bounded)', initMode_B='DataInitMode(Bounded)', initMode_C='DataInitMode(Bounded)', workgroupMappingDim=-1, workgroupMappingValue=-1)GEMMSolution(mac_m=128, mac_n=64, mac_k=16, wave_m=32, wave_n=32, wave_k=8, wave_b=1, workgroup_size_x=128, workgroup_size_y=2, workgroupRemapXCC=False, workgroupRemapXCCValue=-1, unroll_x=0, unroll_y=0, load_A='BufferToLDSViaVGPR', load_B='BufferToLDSViaVGPR', storeLDS_D=True, betaInFma=True, scheduler='Priority', schedulerCost='LinearWeighted', prefetch=False, prefetchInFlight=0, prefetchLDSFactor=0, prefetchMixMemOps=False, loadScale_A='BufferToVGPR', loadScale_B='BufferToVGPR', swizzleScale=False, swizzleTileSize={'m': 0, 'k': 0, 'n': 0, 'l': 0}, prefetchScale=False, streamK=True, numWGs=304, streamKTwoTile=False, streamKTwoTileDPFirst=True, architecture={'ArchString': 'gfx942', 'Xnack': False, 'Sramecc': True}, matchMemoryAccess=True, version='')
-   2.08% | p=5.6994e-05 
	| 3. FloatsGEMM(M: 7680, N: 8448, K: 8448, alpha: 2, beta: 0.5, types: {'type_A': 'half', 'type_B': 'half', 'type_C': 'half', 'type_D': 'half', 'type_acc': 'float', 'trans_A': 'N', 'trans_B': 'T', 'scale_A': 'None', 'scaleType_A': 'None', 'scale_B': 'None', 'scaleType_B': 'None', 'scaleBlockSize': -1, 'scaleSkipPermlane': False, 'scaleShuffleTileA': [], 'scaleShuffleTileB': []}, scaleValue_A: 1, scaleValue_B: 1, initMode_A: DataInitMode(Bounded), initMode_B: DataInitMode(Bounded), initMode_C: DataInitMode(Bounded), workgroupMappingDim: -1, workgroupMappingValue: -1)| GEMM(mac_m: 128, mac_n: 64, mac_k: 16, wave_m: 32, wave_n: 32, wave_k: 8, wave_b: 1, workgroup_size_x: 128, workgroup_size_y: 2, workgroupRemapXCC: False, workgroupRemapXCCValue: -1, unroll_x: 0, unroll_y: 0, load_A: BufferToLDSViaVGPR, load_B: BufferToLDSViaVGPR, storeLDS_D: True, betaInFma: True, scheduler: Priority, schedulerCost: LinearWeighted, prefetch: False, prefetchInFlight: 0, prefetchLDSFactor: 0, prefetchMixMemOps: False, loadScale_A: BufferToVGPR, loadScale_B: BufferToVGPR, swizzleScale: False, swizzleTileSize: {'m': 0, 'k': 0, 'n': 0, 'l': 0}, prefetchScale: False, streamK: True, numWGs: 304, streamKTwoTile: True, streamKTwoTileDPFirst: False, architecture: {'ArchString': 'gfx942', 'Xnack': False, 'Sramecc': True}, matchMemoryAccess: True, version: ) | GEMMProblem(M=7680, N=8448, K=8448, alpha=2, beta=0.5, types={'type_A': 'half', 'type_B': 'half', 'type_C': 'half', 'type_D': 'half', 'type_acc': 'float', 'trans_A': 'N', 'trans_B': 'T', 'scale_A': 'None', 'scaleType_A': 'None', 'scale_B': 'None', 'scaleType_B': 'None', 'scaleBlockSize': -1, 'scaleSkipPermlane': False, 'scaleShuffleTileA': [], 'scaleShuffleTileB': []}, scaleValue_A=1, scaleValue_B=1, initMode_A='DataInitMode(Bounded)', initMode_B='DataInitMode(Bounded)', initMode_C='DataInitMode(Bounded)', workgroupMappingDim=-1, workgroupMappingValue=-1)GEMMSolution(mac_m=128, mac_n=64, mac_k=16, wave_m=32, wave_n=32, wave_k=8, wave_b=1, workgroup_size_x=128, workgroup_size_y=2, workgroupRemapXCC=False, workgroupRemapXCCValue=-1, unroll_x=0, unroll_y=0, load_A='BufferToLDSViaVGPR', load_B='BufferToLDSViaVGPR', storeLDS_D=True, betaInFma=True, scheduler='Priority', schedulerCost='LinearWeighted', prefetch=False, prefetchInFlight=0, prefetchLDSFactor=0, prefetchMixMemOps=False, loadScale_A='BufferToVGPR', loadScale_B='BufferToVGPR', swizzleScale=False, swizzleTileSize={'m': 0, 'k': 0, 'n': 0, 'l': 0}, prefetchScale=False, streamK=True, numWGs=304, streamKTwoTile=True, streamKTwoTileDPFirst=False, architecture={'ArchString': 'gfx942', 'Xnack': False, 'Sramecc': True}, matchMemoryAccess=True, version='')
-   2.55% | p=1.7451e-03 
	| 3. FloatsGEMM(M: 7680, N: 8448, K: 8448, alpha: 2, beta: 0.5, types: {'type_A': 'half', 'type_B': 'half', 'type_C': 'half', 'type_D': 'half', 'type_acc': 'float', 'trans_A': 'N', 'trans_B': 'T', 'scale_A': 'None', 'scaleType_A': 'None', 'scale_B': 'None', 'scaleType_B': 'None', 'scaleBlockSize': -1, 'scaleSkipPermlane': False, 'scaleShuffleTileA': [], 'scaleShuffleTileB': []}, scaleValue_A: 1, scaleValue_B: 1, initMode_A: DataInitMode(Bounded), initMode_B: DataInitMode(Bounded), initMode_C: DataInitMode(Bounded), workgroupMappingDim: -1, workgroupMappingValue: -1)| GEMM(mac_m: 128, mac_n: 64, mac_k: 32, wave_m: 32, wave_n: 32, wave_k: 8, wave_b: 1, workgroup_size_x: 128, workgroup_size_y: 2, workgroupRemapXCC: False, workgroupRemapXCCValue: -1, unroll_x: 0, unroll_y: 0, load_A: BufferToLDSViaVGPR, load_B: BufferToLDSViaVGPR, storeLDS_D: True, betaInFma: True, scheduler: Priority, schedulerCost: LinearWeighted, prefetch: False, prefetchInFlight: 0, prefetchLDSFactor: 0, prefetchMixMemOps: False, loadScale_A: BufferToVGPR, loadScale_B: BufferToVGPR, swizzleScale: False, swizzleTileSize: {'m': 0, 'k': 0, 'n': 0, 'l': 0}, prefetchScale: False, streamK: True, numWGs: 304, streamKTwoTile: False, streamKTwoTileDPFirst: True, architecture: {'ArchString': 'gfx942', 'Xnack': False, 'Sramecc': True}, matchMemoryAccess: True, version: ) | GEMMProblem(M=7680, N=8448, K=8448, alpha=2, beta=0.5, types={'type_A': 'half', 'type_B': 'half', 'type_C': 'half', 'type_D': 'half', 'type_acc': 'float', 'trans_A': 'N', 'trans_B': 'T', 'scale_A': 'None', 'scaleType_A': 'None', 'scale_B': 'None', 'scaleType_B': 'None', 'scaleBlockSize': -1, 'scaleSkipPermlane': False, 'scaleShuffleTileA': [], 'scaleShuffleTileB': []}, scaleValue_A=1, scaleValue_B=1, initMode_A='DataInitMode(Bounded)', initMode_B='DataInitMode(Bounded)', initMode_C='DataInitMode(Bounded)', workgroupMappingDim=-1, workgroupMappingValue=-1)GEMMSolution(mac_m=128, mac_n=64, mac_k=32, wave_m=32, wave_n=32, wave_k=8, wave_b=1, workgroup_size_x=128, workgroup_size_y=2, workgroupRemapXCC=False, workgroupRemapXCCValue=-1, unroll_x=0, unroll_y=0, load_A='BufferToLDSViaVGPR', load_B='BufferToLDSViaVGPR', storeLDS_D=True, betaInFma=True, scheduler='Priority', schedulerCost='LinearWeighted', prefetch=False, prefetchInFlight=0, prefetchLDSFactor=0, prefetchMixMemOps=False, loadScale_A='BufferToVGPR', loadScale_B='BufferToVGPR', swizzleScale=False, swizzleTileSize={'m': 0, 'k': 0, 'n': 0, 'l': 0}, prefetchScale=False, streamK=True, numWGs=304, streamKTwoTile=False, streamKTwoTileDPFirst=True, architecture={'ArchString': 'gfx942', 'Xnack': False, 'Sramecc': True}, matchMemoryAccess=True, version='')
+   2.37% | p=1.7451e-03 
	| 3. FloatsGEMM(M: 7680, N: 8448, K: 8448, alpha: 2, beta: 0.5, types: {'type_A': 'half', 'type_B': 'half', 'type_C': 'half', 'type_D': 'half', 'type_acc': 'float', 'trans_A': 'N', 'trans_B': 'T', 'scale_A': 'None', 'scaleType_A': 'None', 'scale_B': 'None', 'scaleType_B': 'None', 'scaleBlockSize': -1, 'scaleSkipPermlane': False, 'scaleShuffleTileA': [], 'scaleShuffleTileB': []}, scaleValue_A: 1, scaleValue_B: 1, initMode_A: DataInitMode(Bounded), initMode_B: DataInitMode(Bounded), initMode_C: DataInitMode(Bounded), workgroupMappingDim: -1, workgroupMappingValue: -1)| GEMM(mac_m: 128, mac_n: 64, mac_k: 32, wave_m: 32, wave_n: 32, wave_k: 8, wave_b: 1, workgroup_size_x: 128, workgroup_size_y: 2, workgroupRemapXCC: False, workgroupRemapXCCValue: -1, unroll_x: 0, unroll_y: 0, load_A: BufferToLDSViaVGPR, load_B: BufferToLDSViaVGPR, storeLDS_D: True, betaInFma: True, scheduler: Priority, schedulerCost: LinearWeighted, prefetch: False, prefetchInFlight: 0, prefetchLDSFactor: 0, prefetchMixMemOps: False, loadScale_A: BufferToVGPR, loadScale_B: BufferToVGPR, swizzleScale: False, swizzleTileSize: {'m': 0, 'k': 0, 'n': 0, 'l': 0}, prefetchScale: False, streamK: True, numWGs: 304, streamKTwoTile: True, streamKTwoTileDPFirst: False, architecture: {'ArchString': 'gfx942', 'Xnack': False, 'Sramecc': True}, matchMemoryAccess: True, version: ) | GEMMProblem(M=7680, N=8448, K=8448, alpha=2, beta=0.5, types={'type_A': 'half', 'type_B': 'half', 'type_C': 'half', 'type_D': 'half', 'type_acc': 'float', 'trans_A': 'N', 'trans_B': 'T', 'scale_A': 'None', 'scaleType_A': 'None', 'scale_B': 'None', 'scaleType_B': 'None', 'scaleBlockSize': -1, 'scaleSkipPermlane': False, 'scaleShuffleTileA': [], 'scaleShuffleTileB': []}, scaleValue_A=1, scaleValue_B=1, initMode_A='DataInitMode(Bounded)', initMode_B='DataInitMode(Bounded)', initMode_C='DataInitMode(Bounded)', workgroupMappingDim=-1, workgroupMappingValue=-1)GEMMSolution(mac_m=128, mac_n=64, mac_k=32, wave_m=32, wave_n=32, wave_k=8, wave_b=1, workgroup_size_x=128, workgroup_size_y=2, workgroupRemapXCC=False, workgroupRemapXCCValue=-1, unroll_x=0, unroll_y=0, load_A='BufferToLDSViaVGPR', load_B='BufferToLDSViaVGPR', storeLDS_D=True, betaInFma=True, scheduler='Priority', schedulerCost='LinearWeighted', prefetch=False, prefetchInFlight=0, prefetchLDSFactor=0, prefetchMixMemOps=False, loadScale_A='BufferToVGPR', loadScale_B='BufferToVGPR', swizzleScale=False, swizzleTileSize={'m': 0, 'k': 0, 'n': 0, 'l': 0}, prefetchScale=False, streamK=True, numWGs=304, streamKTwoTile=True, streamKTwoTileDPFirst=False, architecture={'ArchString': 'gfx942', 'Xnack': False, 'Sramecc': True}, matchMemoryAccess=True, version='')
+   1.01% | p=1.7451e-03 
	| 3. FloatsGEMM(M: 7680, N: 8448, K: 8448, alpha: 2, beta: 0.5, types: {'type_A': 'half', 'type_B': 'half', 'type_C': 'half', 'type_D': 'half', 'type_acc': 'float', 'trans_A': 'N', 'trans_B': 'T', 'scale_A': 'None', 'scaleType_A': 'None', 'scale_B': 'None', 'scaleType_B': 'None', 'scaleBlockSize': -1, 'scaleSkipPermlane': False, 'scaleShuffleTileA': [], 'scaleShuffleTileB': []}, scaleValue_A: 1, scaleValue_B: 1, initMode_A: DataInitMode(Bounded), initMode_B: DataInitMode(Bounded), initMode_C: DataInitMode(Bounded), workgroupMappingDim: -1, workgroupMappingValue: -1)| GEMM(mac_m: 64, mac_n: 128, mac_k: 16, wave_m: 32, wave_n: 32, wave_k: 8, wave_b: 1, workgroup_size_x: 128, workgroup_size_y: 2, workgroupRemapXCC: False, workgroupRemapXCCValue: -1, unroll_x: 0, unroll_y: 0, load_A: BufferToLDSViaVGPR, load_B: BufferToLDSViaVGPR, storeLDS_D: True, betaInFma: True, scheduler: Priority, schedulerCost: LinearWeighted, prefetch: False, prefetchInFlight: 0, prefetchLDSFactor: 0, prefetchMixMemOps: False, loadScale_A: BufferToVGPR, loadScale_B: BufferToVGPR, swizzleScale: False, swizzleTileSize: {'m': 0, 'k': 0, 'n': 0, 'l': 0}, prefetchScale: False, streamK: True, numWGs: 304, streamKTwoTile: True, streamKTwoTileDPFirst: False, architecture: {'ArchString': 'gfx942', 'Xnack': False, 'Sramecc': True}, matchMemoryAccess: True, version: ) | GEMMProblem(M=7680, N=8448, K=8448, alpha=2, beta=0.5, types={'type_A': 'half', 'type_B': 'half', 'type_C': 'half', 'type_D': 'half', 'type_acc': 'float', 'trans_A': 'N', 'trans_B': 'T', 'scale_A': 'None', 'scaleType_A': 'None', 'scale_B': 'None', 'scaleType_B': 'None', 'scaleBlockSize': -1, 'scaleSkipPermlane': False, 'scaleShuffleTileA': [], 'scaleShuffleTileB': []}, scaleValue_A=1, scaleValue_B=1, initMode_A='DataInitMode(Bounded)', initMode_B='DataInitMode(Bounded)', initMode_C='DataInitMode(Bounded)', workgroupMappingDim=-1, workgroupMappingValue=-1)GEMMSolution(mac_m=64, mac_n=128, mac_k=16, wave_m=32, wave_n=32, wave_k=8, wave_b=1, workgroup_size_x=128, workgroup_size_y=2, workgroupRemapXCC=False, workgroupRemapXCCValue=-1, unroll_x=0, unroll_y=0, load_A='BufferToLDSViaVGPR', load_B='BufferToLDSViaVGPR', storeLDS_D=True, betaInFma=True, scheduler='Priority', schedulerCost='LinearWeighted', prefetch=False, prefetchInFlight=0, prefetchLDSFactor=0, prefetchMixMemOps=False, loadScale_A='BufferToVGPR', loadScale_B='BufferToVGPR', swizzleScale=False, swizzleTileSize={'m': 0, 'k': 0, 'n': 0, 'l': 0}, prefetchScale=False, streamK=True, numWGs=304, streamKTwoTile=True, streamKTwoTileDPFirst=False, architecture={'ArchString': 'gfx942', 'Xnack': False, 'Sramecc': True}, matchMemoryAccess=True, version='')
-   2.36% | p=2.5347e-02 
	| 3. FloatsGEMM(M: 7680, N: 8448, K: 8448, alpha: 2, beta: 0.5, types: {'type_A': 'half', 'type_B': 'half', 'type_C': 'half', 'type_D': 'half', 'type_acc': 'float', 'trans_A': 'N', 'trans_B': 'T', 'scale_A': 'None', 'scaleType_A': 'None', 'scale_B': 'None', 'scaleType_B': 'None', 'scaleBlockSize': -1, 'scaleSkipPermlane': False, 'scaleShuffleTileA': [], 'scaleShuffleTileB': []}, scaleValue_A: 1, scaleValue_B: 1, initMode_A: DataInitMode(Bounded), initMode_B: DataInitMode(Bounded), initMode_C: DataInitMode(Bounded), workgroupMappingDim: -1, workgroupMappingValue: -1)| GEMM(mac_m: 64, mac_n: 128, mac_k: 16, wave_m: 32, wave_n: 32, wave_k: 8, wave_b: 1, workgroup_size_x: 128, workgroup_size_y: 2, workgroupRemapXCC: False, workgroupRemapXCCValue: -1, unroll_x: 0, unroll_y: 0, load_A: BufferToLDSViaVGPR, load_B: BufferToLDSViaVGPR, storeLDS_D: True, betaInFma: True, scheduler: Priority, schedulerCost: LinearWeighted, prefetch: True, prefetchInFlight: 2, prefetchLDSFactor: 2, prefetchMixMemOps: False, loadScale_A: BufferToVGPR, loadScale_B: BufferToVGPR, swizzleScale: False, swizzleTileSize: {'m': 0, 'k': 0, 'n': 0, 'l': 0}, prefetchScale: False, streamK: True, numWGs: 304, streamKTwoTile: False, streamKTwoTileDPFirst: False, architecture: {'ArchString': 'gfx942', 'Xnack': False, 'Sramecc': True}, matchMemoryAccess: True, version: ) | GEMMProblem(M=7680, N=8448, K=8448, alpha=2, beta=0.5, types={'type_A': 'half', 'type_B': 'half', 'type_C': 'half', 'type_D': 'half', 'type_acc': 'float', 'trans_A': 'N', 'trans_B': 'T', 'scale_A': 'None', 'scaleType_A': 'None', 'scale_B': 'None', 'scaleType_B': 'None', 'scaleBlockSize': -1, 'scaleSkipPermlane': False, 'scaleShuffleTileA': [], 'scaleShuffleTileB': []}, scaleValue_A=1, scaleValue_B=1, initMode_A='DataInitMode(Bounded)', initMode_B='DataInitMode(Bounded)', initMode_C='DataInitMode(Bounded)', workgroupMappingDim=-1, workgroupMappingValue=-1)GEMMSolution(mac_m=64, mac_n=128, mac_k=16, wave_m=32, wave_n=32, wave_k=8, wave_b=1, workgroup_size_x=128, workgroup_size_y=2, workgroupRemapXCC=False, workgroupRemapXCCValue=-1, unroll_x=0, unroll_y=0, load_A='BufferToLDSViaVGPR', load_B='BufferToLDSViaVGPR', storeLDS_D=True, betaInFma=True, scheduler='Priority', schedulerCost='LinearWeighted', prefetch=True, prefetchInFlight=2, prefetchLDSFactor=2, prefetchMixMemOps=False, loadScale_A='BufferToVGPR', loadScale_B='BufferToVGPR', swizzleScale=False, swizzleTileSize={'m': 0, 'k': 0, 'n': 0, 'l': 0}, prefetchScale=False, streamK=True, numWGs=304, streamKTwoTile=False, streamKTwoTileDPFirst=False, architecture={'ArchString': 'gfx942', 'Xnack': False, 'Sramecc': True}, matchMemoryAccess=True, version='')
-   0.86% | p=2.5347e-02 
	| 3. FloatsGEMM(M: 7680, N: 8448, K: 8448, alpha: 2, beta: 0.5, types: {'type_A': 'half', 'type_B': 'half', 'type_C': 'half', 'type_D': 'half', 'type_acc': 'float', 'trans_A': 'N', 'trans_B': 'T', 'scale_A': 'None', 'scaleType_A': 'None', 'scale_B': 'None', 'scaleType_B': 'None', 'scaleBlockSize': -1, 'scaleSkipPermlane': False, 'scaleShuffleTileA': [], 'scaleShuffleTileB': []}, scaleValue_A: 1, scaleValue_B: 1, initMode_A: DataInitMode(Bounded), initMode_B: DataInitMode(Bounded), initMode_C: DataInitMode(Bounded), workgroupMappingDim: -1, workgroupMappingValue: -1)| GEMM(mac_m: 64, mac_n: 128, mac_k: 32, wave_m: 32, wave_n: 32, wave_k: 8, wave_b: 1, workgroup_size_x: 128, workgroup_size_y: 2, workgroupRemapXCC: False, workgroupRemapXCCValue: -1, unroll_x: 0, unroll_y: 0, load_A: BufferToLDSViaVGPR, load_B: BufferToLDSViaVGPR, storeLDS_D: True, betaInFma: True, scheduler: Priority, schedulerCost: LinearWeighted, prefetch: False, prefetchInFlight: 0, prefetchLDSFactor: 0, prefetchMixMemOps: False, loadScale_A: BufferToVGPR, loadScale_B: BufferToVGPR, swizzleScale: False, swizzleTileSize: {'m': 0, 'k': 0, 'n': 0, 'l': 0}, prefetchScale: False, streamK: True, numWGs: 304, streamKTwoTile: False, streamKTwoTileDPFirst: True, architecture: {'ArchString': 'gfx942', 'Xnack': False, 'Sramecc': True}, matchMemoryAccess: True, version: ) | GEMMProblem(M=7680, N=8448, K=8448, alpha=2, beta=0.5, types={'type_A': 'half', 'type_B': 'half', 'type_C': 'half', 'type_D': 'half', 'type_acc': 'float', 'trans_A': 'N', 'trans_B': 'T', 'scale_A': 'None', 'scaleType_A': 'None', 'scale_B': 'None', 'scaleType_B': 'None', 'scaleBlockSize': -1, 'scaleSkipPermlane': False, 'scaleShuffleTileA': [], 'scaleShuffleTileB': []}, scaleValue_A=1, scaleValue_B=1, initMode_A='DataInitMode(Bounded)', initMode_B='DataInitMode(Bounded)', initMode_C='DataInitMode(Bounded)', workgroupMappingDim=-1, workgroupMappingValue=-1)GEMMSolution(mac_m=64, mac_n=128, mac_k=32, wave_m=32, wave_n=32, wave_k=8, wave_b=1, workgroup_size_x=128, workgroup_size_y=2, workgroupRemapXCC=False, workgroupRemapXCCValue=-1, unroll_x=0, unroll_y=0, load_A='BufferToLDSViaVGPR', load_B='BufferToLDSViaVGPR', storeLDS_D=True, betaInFma=True, scheduler='Priority', schedulerCost='LinearWeighted', prefetch=False, prefetchInFlight=0, prefetchLDSFactor=0, prefetchMixMemOps=False, loadScale_A='BufferToVGPR', loadScale_B='BufferToVGPR', swizzleScale=False, swizzleTileSize={'m': 0, 'k': 0, 'n': 0, 'l': 0}, prefetchScale=False, streamK=True, numWGs=304, streamKTwoTile=False, streamKTwoTileDPFirst=True, architecture={'ArchString': 'gfx942', 'Xnack': False, 'Sramecc': True}, matchMemoryAccess=True, version='')
+   3.12% | p=2.5347e-02 
	| 3. FloatsGEMM(M: 7680, N: 8448, K: 8448, alpha: 2, beta: 0.5, types: {'type_A': 'half', 'type_B': 'half', 'type_C': 'half', 'type_D': 'half', 'type_acc': 'float', 'trans_A': 'N', 'trans_B': 'T', 'scale_A': 'None', 'scaleType_A': 'None', 'scale_B': 'None', 'scaleType_B': 'None', 'scaleBlockSize': -1, 'scaleSkipPermlane': False, 'scaleShuffleTileA': [], 'scaleShuffleTileB': []}, scaleValue_A: 1, scaleValue_B: 1, initMode_A: DataInitMode(Bounded), initMode_B: DataInitMode(Bounded), initMode_C: DataInitMode(Bounded), workgroupMappingDim: -1, workgroupMappingValue: -1)| GEMM(mac_m: 64, mac_n: 256, mac_k: 16, wave_m: 32, wave_n: 32, wave_k: 8, wave_b: 1, workgroup_size_x: 128, workgroup_size_y: 2, workgroupRemapXCC: False, workgroupRemapXCCValue: -1, unroll_x: 0, unroll_y: 0, load_A: BufferToLDSViaVGPR, load_B: BufferToLDSViaVGPR, storeLDS_D: True, betaInFma: True, scheduler: Priority, schedulerCost: LinearWeighted, prefetch: False, prefetchInFlight: 0, prefetchLDSFactor: 0, prefetchMixMemOps: False, loadScale_A: BufferToVGPR, loadScale_B: BufferToVGPR, swizzleScale: False, swizzleTileSize: {'m': 0, 'k': 0, 'n': 0, 'l': 0}, prefetchScale: False, streamK: True, numWGs: 304, streamKTwoTile: False, streamKTwoTileDPFirst: False, architecture: {'ArchString': 'gfx942', 'Xnack': False, 'Sramecc': True}, matchMemoryAccess: True, version: ) | GEMMProblem(M=7680, N=8448, K=8448, alpha=2, beta=0.5, types={'type_A': 'half', 'type_B': 'half', 'type_C': 'half', 'type_D': 'half', 'type_acc': 'float', 'trans_A': 'N', 'trans_B': 'T', 'scale_A': 'None', 'scaleType_A': 'None', 'scale_B': 'None', 'scaleType_B': 'None', 'scaleBlockSize': -1, 'scaleSkipPermlane': False, 'scaleShuffleTileA': [], 'scaleShuffleTileB': []}, scaleValue_A=1, scaleValue_B=1, initMode_A='DataInitMode(Bounded)', initMode_B='DataInitMode(Bounded)', initMode_C='DataInitMode(Bounded)', workgroupMappingDim=-1, workgroupMappingValue=-1)GEMMSolution(mac_m=64, mac_n=256, mac_k=16, wave_m=32, wave_n=32, wave_k=8, wave_b=1, workgroup_size_x=128, workgroup_size_y=2, workgroupRemapXCC=False, workgroupRemapXCCValue=-1, unroll_x=0, unroll_y=0, load_A='BufferToLDSViaVGPR', load_B='BufferToLDSViaVGPR', storeLDS_D=True, betaInFma=True, scheduler='Priority', schedulerCost='LinearWeighted', prefetch=False, prefetchInFlight=0, prefetchLDSFactor=0, prefetchMixMemOps=False, loadScale_A='BufferToVGPR', loadScale_B='BufferToVGPR', swizzleScale=False, swizzleTileSize={'m': 0, 'k': 0, 'n': 0, 'l': 0}, prefetchScale=False, streamK=True, numWGs=304, streamKTwoTile=False, streamKTwoTileDPFirst=False, architecture={'ArchString': 'gfx942', 'Xnack': False, 'Sramecc': True}, matchMemoryAccess=True, version='')
-   0.99% | p=1.7451e-03 
	| 3. FloatsGEMM(M: 7680, N: 8448, K: 8448, alpha: 2, beta: 0.5, types: {'type_A': 'half', 'type_B': 'half', 'type_C': 'half', 'type_D': 'half', 'type_acc': 'float', 'trans_A': 'N', 'trans_B': 'T', 'scale_A': 'None', 'scaleType_A': 'None', 'scale_B': 'None', 'scaleType_B': 'None', 'scaleBlockSize': -1, 'scaleSkipPermlane': False, 'scaleShuffleTileA': [], 'scaleShuffleTileB': []}, scaleValue_A: 1, scaleValue_B: 1, initMode_A: DataInitMode(Bounded), initMode_B: DataInitMode(Bounded), initMode_C: DataInitMode(Bounded), workgroupMappingDim: -1, workgroupMappingValue: -1)| GEMM(mac_m: 64, mac_n: 256, mac_k: 16, wave_m: 32, wave_n: 32, wave_k: 8, wave_b: 1, workgroup_size_x: 128, workgroup_size_y: 2, workgroupRemapXCC: False, workgroupRemapXCCValue: -1, unroll_x: 0, unroll_y: 0, load_A: BufferToLDSViaVGPR, load_B: BufferToLDSViaVGPR, storeLDS_D: True, betaInFma: True, scheduler: Priority, schedulerCost: LinearWeighted, prefetch: False, prefetchInFlight: 0, prefetchLDSFactor: 0, prefetchMixMemOps: False, loadScale_A: BufferToVGPR, loadScale_B: BufferToVGPR, swizzleScale: False, swizzleTileSize: {'m': 0, 'k': 0, 'n': 0, 'l': 0}, prefetchScale: False, streamK: True, numWGs: 304, streamKTwoTile: True, streamKTwoTileDPFirst: False, architecture: {'ArchString': 'gfx942', 'Xnack': False, 'Sramecc': True}, matchMemoryAccess: True, version: ) | GEMMProblem(M=7680, N=8448, K=8448, alpha=2, beta=0.5, types={'type_A': 'half', 'type_B': 'half', 'type_C': 'half', 'type_D': 'half', 'type_acc': 'float', 'trans_A': 'N', 'trans_B': 'T', 'scale_A': 'None', 'scaleType_A': 'None', 'scale_B': 'None', 'scaleType_B': 'None', 'scaleBlockSize': -1, 'scaleSkipPermlane': False, 'scaleShuffleTileA': [], 'scaleShuffleTileB': []}, scaleValue_A=1, scaleValue_B=1, initMode_A='DataInitMode(Bounded)', initMode_B='DataInitMode(Bounded)', initMode_C='DataInitMode(Bounded)', workgroupMappingDim=-1, workgroupMappingValue=-1)GEMMSolution(mac_m=64, mac_n=256, mac_k=16, wave_m=32, wave_n=32, wave_k=8, wave_b=1, workgroup_size_x=128, workgroup_size_y=2, workgroupRemapXCC=False, workgroupRemapXCCValue=-1, unroll_x=0, unroll_y=0, load_A='BufferToLDSViaVGPR', load_B='BufferToLDSViaVGPR', storeLDS_D=True, betaInFma=True, scheduler='Priority', schedulerCost='LinearWeighted', prefetch=False, prefetchInFlight=0, prefetchLDSFactor=0, prefetchMixMemOps=False, loadScale_A='BufferToVGPR', loadScale_B='BufferToVGPR', swizzleScale=False, swizzleTileSize={'m': 0, 'k': 0, 'n': 0, 'l': 0}, prefetchScale=False, streamK=True, numWGs=304, streamKTwoTile=True, streamKTwoTileDPFirst=False, architecture={'ArchString': 'gfx942', 'Xnack': False, 'Sramecc': True}, matchMemoryAccess=True, version='')
+   1.50% | p=2.5347e-02 
	| 3. FloatsGEMM(M: 7680, N: 8448, K: 8448, alpha: 2, beta: 0.5, types: {'type_A': 'half', 'type_B': 'half', 'type_C': 'half', 'type_D': 'half', 'type_acc': 'float', 'trans_A': 'N', 'trans_B': 'T', 'scale_A': 'None', 'scaleType_A': 'None', 'scale_B': 'None', 'scaleType_B': 'None', 'scaleBlockSize': -1, 'scaleSkipPermlane': False, 'scaleShuffleTileA': [], 'scaleShuffleTileB': []}, scaleValue_A: 1, scaleValue_B: 1, initMode_A: DataInitMode(Bounded), initMode_B: DataInitMode(Bounded), initMode_C: DataInitMode(Bounded), workgroupMappingDim: -1, workgroupMappingValue: -1)| GEMM(mac_m: 64, mac_n: 256, mac_k: 32, wave_m: 32, wave_n: 32, wave_k: 8, wave_b: 1, workgroup_size_x: 128, workgroup_size_y: 2, workgroupRemapXCC: False, workgroupRemapXCCValue: -1, unroll_x: 0, unroll_y: 0, load_A: BufferToLDSViaVGPR, load_B: BufferToLDSViaVGPR, storeLDS_D: True, betaInFma: True, scheduler: Priority, schedulerCost: LinearWeighted, prefetch: False, prefetchInFlight: 0, prefetchLDSFactor: 0, prefetchMixMemOps: False, loadScale_A: BufferToVGPR, loadScale_B: BufferToVGPR, swizzleScale: False, swizzleTileSize: {'m': 0, 'k': 0, 'n': 0, 'l': 0}, prefetchScale: False, streamK: True, numWGs: 304, streamKTwoTile: False, streamKTwoTileDPFirst: False, architecture: {'ArchString': 'gfx942', 'Xnack': False, 'Sramecc': True}, matchMemoryAccess: True, version: ) | GEMMProblem(M=7680, N=8448, K=8448, alpha=2, beta=0.5, types={'type_A': 'half', 'type_B': 'half', 'type_C': 'half', 'type_D': 'half', 'type_acc': 'float', 'trans_A': 'N', 'trans_B': 'T', 'scale_A': 'None', 'scaleType_A': 'None', 'scale_B': 'None', 'scaleType_B': 'None', 'scaleBlockSize': -1, 'scaleSkipPermlane': False, 'scaleShuffleTileA': [], 'scaleShuffleTileB': []}, scaleValue_A=1, scaleValue_B=1, initMode_A='DataInitMode(Bounded)', initMode_B='DataInitMode(Bounded)', initMode_C='DataInitMode(Bounded)', workgroupMappingDim=-1, workgroupMappingValue=-1)GEMMSolution(mac_m=64, mac_n=256, mac_k=32, wave_m=32, wave_n=32, wave_k=8, wave_b=1, workgroup_size_x=128, workgroup_size_y=2, workgroupRemapXCC=False, workgroupRemapXCCValue=-1, unroll_x=0, unroll_y=0, load_A='BufferToLDSViaVGPR', load_B='BufferToLDSViaVGPR', storeLDS_D=True, betaInFma=True, scheduler='Priority', schedulerCost='LinearWeighted', prefetch=False, prefetchInFlight=0, prefetchLDSFactor=0, prefetchMixMemOps=False, loadScale_A='BufferToVGPR', loadScale_B='BufferToVGPR', swizzleScale=False, swizzleTileSize={'m': 0, 'k': 0, 'n': 0, 'l': 0}, prefetchScale=False, streamK=True, numWGs=304, streamKTwoTile=False, streamKTwoTileDPFirst=False, architecture={'ArchString': 'gfx942', 'Xnack': False, 'Sramecc': True}, matchMemoryAccess=True, version='')
-   0.89% | p=5.6994e-05 
	| 3. FloatsGEMM(M: 7680, N: 8448, K: 8448, alpha: 2, beta: 0.5, types: {'type_A': 'half', 'type_B': 'half', 'type_C': 'half', 'type_D': 'half', 'type_acc': 'float', 'trans_A': 'N', 'trans_B': 'T', 'scale_A': 'None', 'scaleType_A': 'None', 'scale_B': 'None', 'scaleType_B': 'None', 'scaleBlockSize': -1, 'scaleSkipPermlane': False, 'scaleShuffleTileA': [], 'scaleShuffleTileB': []}, scaleValue_A: 1, scaleValue_B: 1, initMode_A: DataInitMode(Bounded), initMode_B: DataInitMode(Bounded), initMode_C: DataInitMode(Bounded), workgroupMappingDim: -1, workgroupMappingValue: -1)| GEMM(mac_m: 64, mac_n: 64, mac_k: 16, wave_m: 32, wave_n: 32, wave_k: 8, wave_b: 1, workgroup_size_x: 128, workgroup_size_y: 2, workgroupRemapXCC: False, workgroupRemapXCCValue: -1, unroll_x: 0, unroll_y: 0, load_A: BufferToLDSViaVGPR, load_B: BufferToLDSViaVGPR, storeLDS_D: True, betaInFma: True, scheduler: Priority, schedulerCost: LinearWeighted, prefetch: False, prefetchInFlight: 0, prefetchLDSFactor: 0, prefetchMixMemOps: False, loadScale_A: BufferToVGPR, loadScale_B: BufferToVGPR, swizzleScale: False, swizzleTileSize: {'m': 0, 'k': 0, 'n': 0, 'l': 0}, prefetchScale: False, streamK: True, numWGs: 304, streamKTwoTile: False, streamKTwoTileDPFirst: False, architecture: {'ArchString': 'gfx942', 'Xnack': False, 'Sramecc': True}, matchMemoryAccess: True, version: ) | GEMMProblem(M=7680, N=8448, K=8448, alpha=2, beta=0.5, types={'type_A': 'half', 'type_B': 'half', 'type_C': 'half', 'type_D': 'half', 'type_acc': 'float', 'trans_A': 'N', 'trans_B': 'T', 'scale_A': 'None', 'scaleType_A': 'None', 'scale_B': 'None', 'scaleType_B': 'None', 'scaleBlockSize': -1, 'scaleSkipPermlane': False, 'scaleShuffleTileA': [], 'scaleShuffleTileB': []}, scaleValue_A=1, scaleValue_B=1, initMode_A='DataInitMode(Bounded)', initMode_B='DataInitMode(Bounded)', initMode_C='DataInitMode(Bounded)', workgroupMappingDim=-1, workgroupMappingValue=-1)GEMMSolution(mac_m=64, mac_n=64, mac_k=16, wave_m=32, wave_n=32, wave_k=8, wave_b=1, workgroup_size_x=128, workgroup_size_y=2, workgroupRemapXCC=False, workgroupRemapXCCValue=-1, unroll_x=0, unroll_y=0, load_A='BufferToLDSViaVGPR', load_B='BufferToLDSViaVGPR', storeLDS_D=True, betaInFma=True, scheduler='Priority', schedulerCost='LinearWeighted', prefetch=False, prefetchInFlight=0, prefetchLDSFactor=0, prefetchMixMemOps=False, loadScale_A='BufferToVGPR', loadScale_B='BufferToVGPR', swizzleScale=False, swizzleTileSize={'m': 0, 'k': 0, 'n': 0, 'l': 0}, prefetchScale=False, streamK=True, numWGs=304, streamKTwoTile=False, streamKTwoTileDPFirst=False, architecture={'ArchString': 'gfx942', 'Xnack': False, 'Sramecc': True}, matchMemoryAccess=True, version='')
+   1.18% | p=5.6994e-05 
	| 3. FloatsGEMM(M: 7680, N: 8448, K: 8448, alpha: 2, beta: 0.5, types: {'type_A': 'half', 'type_B': 'half', 'type_C': 'half', 'type_D': 'half', 'type_acc': 'float', 'trans_A': 'N', 'trans_B': 'T', 'scale_A': 'None', 'scaleType_A': 'None', 'scale_B': 'None', 'scaleType_B': 'None', 'scaleBlockSize': -1, 'scaleSkipPermlane': False, 'scaleShuffleTileA': [], 'scaleShuffleTileB': []}, scaleValue_A: 1, scaleValue_B: 1, initMode_A: DataInitMode(Bounded), initMode_B: DataInitMode(Bounded), initMode_C: DataInitMode(Bounded), workgroupMappingDim: -1, workgroupMappingValue: -1)| GEMM(mac_m: 64, mac_n: 64, mac_k: 16, wave_m: 32, wave_n: 32, wave_k: 8, wave_b: 1, workgroup_size_x: 128, workgroup_size_y: 2, workgroupRemapXCC: False, workgroupRemapXCCValue: -1, unroll_x: 0, unroll_y: 0, load_A: BufferToLDSViaVGPR, load_B: BufferToLDSViaVGPR, storeLDS_D: True, betaInFma: True, scheduler: Priority, schedulerCost: LinearWeighted, prefetch: False, prefetchInFlight: 0, prefetchLDSFactor: 0, prefetchMixMemOps: False, loadScale_A: BufferToVGPR, loadScale_B: BufferToVGPR, swizzleScale: False, swizzleTileSize: {'m': 0, 'k': 0, 'n': 0, 'l': 0}, prefetchScale: False, streamK: True, numWGs: 304, streamKTwoTile: True, streamKTwoTileDPFirst: False, architecture: {'ArchString': 'gfx942', 'Xnack': False, 'Sramecc': True}, matchMemoryAccess: True, version: ) | GEMMProblem(M=7680, N=8448, K=8448, alpha=2, beta=0.5, types={'type_A': 'half', 'type_B': 'half', 'type_C': 'half', 'type_D': 'half', 'type_acc': 'float', 'trans_A': 'N', 'trans_B': 'T', 'scale_A': 'None', 'scaleType_A': 'None', 'scale_B': 'None', 'scaleType_B': 'None', 'scaleBlockSize': -1, 'scaleSkipPermlane': False, 'scaleShuffleTileA': [], 'scaleShuffleTileB': []}, scaleValue_A=1, scaleValue_B=1, initMode_A='DataInitMode(Bounded)', initMode_B='DataInitMode(Bounded)', initMode_C='DataInitMode(Bounded)', workgroupMappingDim=-1, workgroupMappingValue=-1)GEMMSolution(mac_m=64, mac_n=64, mac_k=16, wave_m=32, wave_n=32, wave_k=8, wave_b=1, workgroup_size_x=128, workgroup_size_y=2, workgroupRemapXCC=False, workgroupRemapXCCValue=-1, unroll_x=0, unroll_y=0, load_A='BufferToLDSViaVGPR', load_B='BufferToLDSViaVGPR', storeLDS_D=True, betaInFma=True, scheduler='Priority', schedulerCost='LinearWeighted', prefetch=False, prefetchInFlight=0, prefetchLDSFactor=0, prefetchMixMemOps=False, loadScale_A='BufferToVGPR', loadScale_B='BufferToVGPR', swizzleScale=False, swizzleTileSize={'m': 0, 'k': 0, 'n': 0, 'l': 0}, prefetchScale=False, streamK=True, numWGs=304, streamKTwoTile=True, streamKTwoTileDPFirst=False, architecture={'ArchString': 'gfx942', 'Xnack': False, 'Sramecc': True}, matchMemoryAccess=True, version='')
+   2.23% | p=5.6994e-05 
	| 3. FloatsGEMM(M: 7680, N: 8448, K: 8448, alpha: 2, beta: 0.5, types: {'type_A': 'half', 'type_B': 'half', 'type_C': 'half', 'type_D': 'half', 'type_acc': 'float', 'trans_A': 'N', 'trans_B': 'T', 'scale_A': 'None', 'scaleType_A': 'None', 'scale_B': 'None', 'scaleType_B': 'None', 'scaleBlockSize': -1, 'scaleSkipPermlane': False, 'scaleShuffleTileA': [], 'scaleShuffleTileB': []}, scaleValue_A: 1, scaleValue_B: 1, initMode_A: DataInitMode(Bounded), initMode_B: DataInitMode(Bounded), initMode_C: DataInitMode(Bounded), workgroupMappingDim: -1, workgroupMappingValue: -1)| GEMM(mac_m: 64, mac_n: 64, mac_k: 32, wave_m: 32, wave_n: 32, wave_k: 8, wave_b: 1, workgroup_size_x: 128, workgroup_size_y: 2, workgroupRemapXCC: False, workgroupRemapXCCValue: -1, unroll_x: 0, unroll_y: 0, load_A: BufferToLDSViaVGPR, load_B: BufferToLDSViaVGPR, storeLDS_D: True, betaInFma: True, scheduler: Priority, schedulerCost: LinearWeighted, prefetch: False, prefetchInFlight: 0, prefetchLDSFactor: 0, prefetchMixMemOps: False, loadScale_A: BufferToVGPR, loadScale_B: BufferToVGPR, swizzleScale: False, swizzleTileSize: {'m': 0, 'k': 0, 'n': 0, 'l': 0}, prefetchScale: False, streamK: True, numWGs: 304, streamKTwoTile: False, streamKTwoTileDPFirst: False, architecture: {'ArchString': 'gfx942', 'Xnack': False, 'Sramecc': True}, matchMemoryAccess: True, version: ) | GEMMProblem(M=7680, N=8448, K=8448, alpha=2, beta=0.5, types={'type_A': 'half', 'type_B': 'half', 'type_C': 'half', 'type_D': 'half', 'type_acc': 'float', 'trans_A': 'N', 'trans_B': 'T', 'scale_A': 'None', 'scaleType_A': 'None', 'scale_B': 'None', 'scaleType_B': 'None', 'scaleBlockSize': -1, 'scaleSkipPermlane': False, 'scaleShuffleTileA': [], 'scaleShuffleTileB': []}, scaleValue_A=1, scaleValue_B=1, initMode_A='DataInitMode(Bounded)', initMode_B='DataInitMode(Bounded)', initMode_C='DataInitMode(Bounded)', workgroupMappingDim=-1, workgroupMappingValue=-1)GEMMSolution(mac_m=64, mac_n=64, mac_k=32, wave_m=32, wave_n=32, wave_k=8, wave_b=1, workgroup_size_x=128, workgroup_size_y=2, workgroupRemapXCC=False, workgroupRemapXCCValue=-1, unroll_x=0, unroll_y=0, load_A='BufferToLDSViaVGPR', load_B='BufferToLDSViaVGPR', storeLDS_D=True, betaInFma=True, scheduler='Priority', schedulerCost='LinearWeighted', prefetch=False, prefetchInFlight=0, prefetchLDSFactor=0, prefetchMixMemOps=False, loadScale_A='BufferToVGPR', loadScale_B='BufferToVGPR', swizzleScale=False, swizzleTileSize={'m': 0, 'k': 0, 'n': 0, 'l': 0}, prefetchScale=False, streamK=True, numWGs=304, streamKTwoTile=False, streamKTwoTileDPFirst=False, architecture={'ArchString': 'gfx942', 'Xnack': False, 'Sramecc': True}, matchMemoryAccess=True, version='')
+   1.79% | p=5.6994e-05 
	| 3. FloatsGEMM(M: 7680, N: 8448, K: 8448, alpha: 2, beta: 0.5, types: {'type_A': 'half', 'type_B': 'half', 'type_C': 'half', 'type_D': 'half', 'type_acc': 'float', 'trans_A': 'N', 'trans_B': 'T', 'scale_A': 'None', 'scaleType_A': 'None', 'scale_B': 'None', 'scaleType_B': 'None', 'scaleBlockSize': -1, 'scaleSkipPermlane': False, 'scaleShuffleTileA': [], 'scaleShuffleTileB': []}, scaleValue_A: 1, scaleValue_B: 1, initMode_A: DataInitMode(Bounded), initMode_B: DataInitMode(Bounded), initMode_C: DataInitMode(Bounded), workgroupMappingDim: -1, workgroupMappingValue: -1)| GEMM(mac_m: 64, mac_n: 64, mac_k: 64, wave_m: 32, wave_n: 32, wave_k: 8, wave_b: 1, workgroup_size_x: 128, workgroup_size_y: 2, workgroupRemapXCC: False, workgroupRemapXCCValue: -1, unroll_x: 0, unroll_y: 0, load_A: BufferToLDSViaVGPR, load_B: BufferToLDSViaVGPR, storeLDS_D: True, betaInFma: True, scheduler: Priority, schedulerCost: LinearWeighted, prefetch: False, prefetchInFlight: 0, prefetchLDSFactor: 0, prefetchMixMemOps: False, loadScale_A: BufferToVGPR, loadScale_B: BufferToVGPR, swizzleScale: False, swizzleTileSize: {'m': 0, 'k': 0, 'n': 0, 'l': 0}, prefetchScale: False, streamK: True, numWGs: 304, streamKTwoTile: False, streamKTwoTileDPFirst: False, architecture: {'ArchString': 'gfx942', 'Xnack': False, 'Sramecc': True}, matchMemoryAccess: True, version: ) | GEMMProblem(M=7680, N=8448, K=8448, alpha=2, beta=0.5, types={'type_A': 'half', 'type_B': 'half', 'type_C': 'half', 'type_D': 'half', 'type_acc': 'float', 'trans_A': 'N', 'trans_B': 'T', 'scale_A': 'None', 'scaleType_A': 'None', 'scale_B': 'None', 'scaleType_B': 'None', 'scaleBlockSize': -1, 'scaleSkipPermlane': False, 'scaleShuffleTileA': [], 'scaleShuffleTileB': []}, scaleValue_A=1, scaleValue_B=1, initMode_A='DataInitMode(Bounded)', initMode_B='DataInitMode(Bounded)', initMode_C='DataInitMode(Bounded)', workgroupMappingDim=-1, workgroupMappingValue=-1)GEMMSolution(mac_m=64, mac_n=64, mac_k=64, wave_m=32, wave_n=32, wave_k=8, wave_b=1, workgroup_size_x=128, workgroup_size_y=2, workgroupRemapXCC=False, workgroupRemapXCCValue=-1, unroll_x=0, unroll_y=0, load_A='BufferToLDSViaVGPR', load_B='BufferToLDSViaVGPR', storeLDS_D=True, betaInFma=True, scheduler='Priority', schedulerCost='LinearWeighted', prefetch=False, prefetchInFlight=0, prefetchLDSFactor=0, prefetchMixMemOps=False, loadScale_A='BufferToVGPR', loadScale_B='BufferToVGPR', swizzleScale=False, swizzleTileSize={'m': 0, 'k': 0, 'n': 0, 'l': 0}, prefetchScale=False, streamK=True, numWGs=304, streamKTwoTile=False, streamKTwoTileDPFirst=False, architecture={'ArchString': 'gfx942', 'Xnack': False, 'Sramecc': True}, matchMemoryAccess=True, version='')
-   1.70% | p=5.6994e-05 
	| 3. FloatsGEMM(M: 7680, N: 8448, K: 8448, alpha: 2, beta: 0.5, types: {'type_A': 'half', 'type_B': 'half', 'type_C': 'half', 'type_D': 'half', 'type_acc': 'float', 'trans_A': 'N', 'trans_B': 'T', 'scale_A': 'None', 'scaleType_A': 'None', 'scale_B': 'None', 'scaleType_B': 'None', 'scaleBlockSize': -1, 'scaleSkipPermlane': False, 'scaleShuffleTileA': [], 'scaleShuffleTileB': []}, scaleValue_A: 1, scaleValue_B: 1, initMode_A: DataInitMode(Bounded), initMode_B: DataInitMode(Bounded), initMode_C: DataInitMode(Bounded), workgroupMappingDim: -1, workgroupMappingValue: -1)| GEMM(mac_m: 64, mac_n: 64, mac_k: 64, wave_m: 32, wave_n: 32, wave_k: 8, wave_b: 1, workgroup_size_x: 128, workgroup_size_y: 2, workgroupRemapXCC: False, workgroupRemapXCCValue: -1, unroll_x: 0, unroll_y: 0, load_A: BufferToLDSViaVGPR, load_B: BufferToLDSViaVGPR, storeLDS_D: True, betaInFma: True, scheduler: Priority, schedulerCost: LinearWeighted, prefetch: False, prefetchInFlight: 0, prefetchLDSFactor: 0, prefetchMixMemOps: False, loadScale_A: BufferToVGPR, loadScale_B: BufferToVGPR, swizzleScale: False, swizzleTileSize: {'m': 0, 'k': 0, 'n': 0, 'l': 0}, prefetchScale: False, streamK: True, numWGs: 304, streamKTwoTile: False, streamKTwoTileDPFirst: True, architecture: {'ArchString': 'gfx942', 'Xnack': False, 'Sramecc': True}, matchMemoryAccess: True, version: ) | GEMMProblem(M=7680, N=8448, K=8448, alpha=2, beta=0.5, types={'type_A': 'half', 'type_B': 'half', 'type_C': 'half', 'type_D': 'half', 'type_acc': 'float', 'trans_A': 'N', 'trans_B': 'T', 'scale_A': 'None', 'scaleType_A': 'None', 'scale_B': 'None', 'scaleType_B': 'None', 'scaleBlockSize': -1, 'scaleSkipPermlane': False, 'scaleShuffleTileA': [], 'scaleShuffleTileB': []}, scaleValue_A=1, scaleValue_B=1, initMode_A='DataInitMode(Bounded)', initMode_B='DataInitMode(Bounded)', initMode_C='DataInitMode(Bounded)', workgroupMappingDim=-1, workgroupMappingValue=-1)GEMMSolution(mac_m=64, mac_n=64, mac_k=64, wave_m=32, wave_n=32, wave_k=8, wave_b=1, workgroup_size_x=128, workgroup_size_y=2, workgroupRemapXCC=False, workgroupRemapXCCValue=-1, unroll_x=0, unroll_y=0, load_A='BufferToLDSViaVGPR', load_B='BufferToLDSViaVGPR', storeLDS_D=True, betaInFma=True, scheduler='Priority', schedulerCost='LinearWeighted', prefetch=False, prefetchInFlight=0, prefetchLDSFactor=0, prefetchMixMemOps=False, loadScale_A='BufferToVGPR', loadScale_B='BufferToVGPR', swizzleScale=False, swizzleTileSize={'m': 0, 'k': 0, 'n': 0, 'l': 0}, prefetchScale=False, streamK=True, numWGs=304, streamKTwoTile=False, streamKTwoTileDPFirst=True, architecture={'ArchString': 'gfx942', 'Xnack': False, 'Sramecc': True}, matchMemoryAccess=True, version='')
+   4.94% | p=5.6994e-05 
	| 3. FloatsGEMM(M: 8448, N: 8448, K: 128, alpha: 2, beta: 0.5, types: {'type_A': 'half', 'type_B': 'half', 'type_C': 'half', 'type_D': 'half', 'type_acc': 'float', 'trans_A': 'N', 'trans_B': 'T', 'scale_A': 'None', 'scaleType_A': 'None', 'scale_B': 'None', 'scaleType_B': 'None', 'scaleBlockSize': -1, 'scaleSkipPermlane': False, 'scaleShuffleTileA': [], 'scaleShuffleTileB': []}, scaleValue_A: 1, scaleValue_B: 1, initMode_A: DataInitMode(Bounded), initMode_B: DataInitMode(Bounded), initMode_C: DataInitMode(Bounded), workgroupMappingDim: -1, workgroupMappingValue: -1)| GEMM(mac_m: 128, mac_n: 256, mac_k: 16, wave_m: 32, wave_n: 32, wave_k: 8, wave_b: 1, workgroup_size_x: 128, workgroup_size_y: 2, workgroupRemapXCC: False, workgroupRemapXCCValue: -1, unroll_x: 0, unroll_y: 0, load_A: BufferToLDSViaVGPR, load_B: BufferToLDSViaVGPR, storeLDS_D: True, betaInFma: True, scheduler: Cooperative, schedulerCost: LinearWeighted, prefetch: True, prefetchInFlight: 2, prefetchLDSFactor: 0, prefetchMixMemOps: False, loadScale_A: BufferToVGPR, loadScale_B: BufferToVGPR, swizzleScale: False, swizzleTileSize: {'m': 0, 'k': 0, 'n': 0, 'l': 0}, prefetchScale: False, streamK: False, numWGs: 0, streamKTwoTile: False, streamKTwoTileDPFirst: False, architecture: {'ArchString': 'gfx942', 'Xnack': False, 'Sramecc': True}, matchMemoryAccess: True, version: ) | GEMMProblem(M=8448, N=8448, K=128, alpha=2, beta=0.5, types={'type_A': 'half', 'type_B': 'half', 'type_C': 'half', 'type_D': 'half', 'type_acc': 'float', 'trans_A': 'N', 'trans_B': 'T', 'scale_A': 'None', 'scaleType_A': 'None', 'scale_B': 'None', 'scaleType_B': 'None', 'scaleBlockSize': -1, 'scaleSkipPermlane': False, 'scaleShuffleTileA': [], 'scaleShuffleTileB': []}, scaleValue_A=1, scaleValue_B=1, initMode_A='DataInitMode(Bounded)', initMode_B='DataInitMode(Bounded)', initMode_C='DataInitMode(Bounded)', workgroupMappingDim=-1, workgroupMappingValue=-1)GEMMSolution(mac_m=128, mac_n=256, mac_k=16, wave_m=32, wave_n=32, wave_k=8, wave_b=1, workgroup_size_x=128, workgroup_size_y=2, workgroupRemapXCC=False, workgroupRemapXCCValue=-1, unroll_x=0, unroll_y=0, load_A='BufferToLDSViaVGPR', load_B='BufferToLDSViaVGPR', storeLDS_D=True, betaInFma=True, scheduler='Cooperative', schedulerCost='LinearWeighted', prefetch=True, prefetchInFlight=2, prefetchLDSFactor=0, prefetchMixMemOps=False, loadScale_A='BufferToVGPR', loadScale_B='BufferToVGPR', swizzleScale=False, swizzleTileSize={'m': 0, 'k': 0, 'n': 0, 'l': 0}, prefetchScale=False, streamK=False, numWGs=0, streamKTwoTile=False, streamKTwoTileDPFirst=False, architecture={'ArchString': 'gfx942', 'Xnack': False, 'Sramecc': True}, matchMemoryAccess=True, version='')
-   7.78% | p=5.6994e-05 
	| 3. FloatsGEMM(M: 8448, N: 8448, K: 128, alpha: 2, beta: 0.5, types: {'type_A': 'half', 'type_B': 'half', 'type_C': 'half', 'type_D': 'half', 'type_acc': 'float', 'trans_A': 'N', 'trans_B': 'T', 'scale_A': 'None', 'scaleType_A': 'None', 'scale_B': 'None', 'scaleType_B': 'None', 'scaleBlockSize': -1, 'scaleSkipPermlane': False, 'scaleShuffleTileA': [], 'scaleShuffleTileB': []}, scaleValue_A: 1, scaleValue_B: 1, initMode_A: DataInitMode(Bounded), initMode_B: DataInitMode(Bounded), initMode_C: DataInitMode(Bounded), workgroupMappingDim: -1, workgroupMappingValue: -1)| GEMM(mac_m: 128, mac_n: 256, mac_k: 16, wave_m: 32, wave_n: 32, wave_k: 8, wave_b: 1, workgroup_size_x: 128, workgroup_size_y: 2, workgroupRemapXCC: False, workgroupRemapXCCValue: -1, unroll_x: 0, unroll_y: 0, load_A: BufferToLDSViaVGPR, load_B: BufferToLDSViaVGPR, storeLDS_D: True, betaInFma: True, scheduler: Priority, schedulerCost: LinearWeighted, prefetch: True, prefetchInFlight: 2, prefetchLDSFactor: 0, prefetchMixMemOps: False, loadScale_A: BufferToVGPR, loadScale_B: BufferToVGPR, swizzleScale: False, swizzleTileSize: {'m': 0, 'k': 0, 'n': 0, 'l': 0}, prefetchScale: False, streamK: False, numWGs: 0, streamKTwoTile: False, streamKTwoTileDPFirst: False, architecture: {'ArchString': 'gfx942', 'Xnack': False, 'Sramecc': True}, matchMemoryAccess: True, version: ) | GEMMProblem(M=8448, N=8448, K=128, alpha=2, beta=0.5, types={'type_A': 'half', 'type_B': 'half', 'type_C': 'half', 'type_D': 'half', 'type_acc': 'float', 'trans_A': 'N', 'trans_B': 'T', 'scale_A': 'None', 'scaleType_A': 'None', 'scale_B': 'None', 'scaleType_B': 'None', 'scaleBlockSize': -1, 'scaleSkipPermlane': False, 'scaleShuffleTileA': [], 'scaleShuffleTileB': []}, scaleValue_A=1, scaleValue_B=1, initMode_A='DataInitMode(Bounded)', initMode_B='DataInitMode(Bounded)', initMode_C='DataInitMode(Bounded)', workgroupMappingDim=-1, workgroupMappingValue=-1)GEMMSolution(mac_m=128, mac_n=256, mac_k=16, wave_m=32, wave_n=32, wave_k=8, wave_b=1, workgroup_size_x=128, workgroup_size_y=2, workgroupRemapXCC=False, workgroupRemapXCCValue=-1, unroll_x=0, unroll_y=0, load_A='BufferToLDSViaVGPR', load_B='BufferToLDSViaVGPR', storeLDS_D=True, betaInFma=True, scheduler='Priority', schedulerCost='LinearWeighted', prefetch=True, prefetchInFlight=2, prefetchLDSFactor=0, prefetchMixMemOps=False, loadScale_A='BufferToVGPR', loadScale_B='BufferToVGPR', swizzleScale=False, swizzleTileSize={'m': 0, 'k': 0, 'n': 0, 'l': 0}, prefetchScale=False, streamK=False, numWGs=0, streamKTwoTile=False, streamKTwoTileDPFirst=False, architecture={'ArchString': 'gfx942', 'Xnack': False, 'Sramecc': True}, matchMemoryAccess=True, version='')
+   0.86% | p=2.2479e-02 
	| CodeGen(instCount: 40000, instructions: comments)| CodeGen() | CodeGen(instCount: 40000, instructions: comments)
-   0.68% | p=2.5347e-02 
	| CodeGen(instCount: 40000, instructions: complex_mi_with_coop)| CodeGen() | CodeGen(instCount: 40000, instructions: complex_mi_with_coop)
+   1.39% | p=1.7451e-03 
	| CodeGen(instCount: 40000, instructions: simple_mi)| CodeGen() | CodeGen(instCount: 40000, instructions: simple_mi)

Links

math-ci · 2025-12-06T03:09:10Z

Resource Report for gfx942

Results

Details

@@                    Resource Usage Changes                         @@
====================================================================================================
- SGPR: 68 -> 70 (+2)
	| 3. FloatsGEMM(M: 3072, N: 4096, K: 4096, alpha: 2, beta: 0.5, types: {'type_A': 'float', 'type_B': 'float', 'type_C': 'float', 'type_D': 'float', 'type_acc': 'float', 'trans_A': 'N', 'trans_B': 'T', 'scale_A': 'None', 'scaleType_A': 'None', 'scale_B': 'None', 'scaleType_B': 'None', 'scaleBlockSize': -1, 'scaleSkipPermlane': False, 'scaleShuffleTileA': [], 'scaleShuffleTileB': []}, scaleValue_A: 1, scaleValue_B: 1, initMode_A: DataInitMode(Bounded), initMode_B: DataInitMode(Bounded), initMode_C: DataInitMode(Bounded), workgroupMappingDim: -1, workgroupMappingValue: -1)| GEMM(mac_m: 64, mac_n: 64, mac_k: 64, wave_m: 32, wave_n: 32, wave_k: 2, wave_b: 1, workgroup_size_x: 128, workgroup_size_y: 2, workgroupRemapXCC: False, workgroupRemapXCCValue: -1, unroll_x: 0, unroll_y: 0, load_A: BufferToLDSViaVGPR, load_B: BufferToLDSViaVGPR, storeLDS_D: True, betaInFma: True, scheduler: Priority, schedulerCost: LinearWeighted, prefetch: False, prefetchInFlight: 2, prefetchLDSFactor: 0, prefetchMixMemOps: False, loadScale_A: BufferToVGPR, loadScale_B: BufferToVGPR, swizzleScale: False, swizzleTileSize: {'m': 0, 'k': 0, 'n': 0, 'l': 0}, prefetchScale: False, streamK: True, numWGs: 304, streamKTwoTile: False, streamKTwoTileDPFirst: False, architecture: {'ArchString': 'gfx942', 'Xnack': False, 'Sramecc': True}, matchMemoryAccess: True, version: ) | GEMMProblem(M=3072, N=4096, K=4096, alpha=2, beta=0.5, types={'type_A': 'float', 'type_B': 'float', 'type_C': 'float', 'type_D': 'float', 'type_acc': 'float', 'trans_A': 'N', 'trans_B': 'T', 'scale_A': 'None', 'scaleType_A': 'None', 'scale_B': 'None', 'scaleType_B': 'None', 'scaleBlockSize': -1, 'scaleSkipPermlane': False, 'scaleShuffleTileA': [], 'scaleShuffleTileB': []}, scaleValue_A=1, scaleValue_B=1, initMode_A='DataInitMode(Bounded)', initMode_B='DataInitMode(Bounded)', initMode_C='DataInitMode(Bounded)', workgroupMappingDim=-1, workgroupMappingValue=-1)GEMMSolution(mac_m=64, mac_n=64, mac_k=64, wave_m=32, wave_n=32, wave_k=2, wave_b=1, workgroup_size_x=128, workgroup_size_y=2, workgroupRemapXCC=False, workgroupRemapXCCValue=-1, unroll_x=0, unroll_y=0, load_A='BufferToLDSViaVGPR', load_B='BufferToLDSViaVGPR', storeLDS_D=True, betaInFma=True, scheduler='Priority', schedulerCost='LinearWeighted', prefetch=False, prefetchInFlight=2, prefetchLDSFactor=0, prefetchMixMemOps=False, loadScale_A='BufferToVGPR', loadScale_B='BufferToVGPR', swizzleScale=False, swizzleTileSize={'m': 0, 'k': 0, 'n': 0, 'l': 0}, prefetchScale=False, streamK=True, numWGs=304, streamKTwoTile=False, streamKTwoTileDPFirst=False, architecture={'ArchString': 'gfx942', 'Xnack': False, 'Sramecc': True}, matchMemoryAccess=True, version='')
- SGPR: 72 -> 74 (+2)
	| 3. FloatsGEMM(M: 3072, N: 4096, K: 4096, alpha: 2, beta: 0.5, types: {'type_A': 'float', 'type_B': 'float', 'type_C': 'float', 'type_D': 'float', 'type_acc': 'float', 'trans_A': 'N', 'trans_B': 'T', 'scale_A': 'None', 'scaleType_A': 'None', 'scale_B': 'None', 'scaleType_B': 'None', 'scaleBlockSize': -1, 'scaleSkipPermlane': False, 'scaleShuffleTileA': [], 'scaleShuffleTileB': []}, scaleValue_A: 1, scaleValue_B: 1, initMode_A: DataInitMode(Bounded), initMode_B: DataInitMode(Bounded), initMode_C: DataInitMode(Bounded), workgroupMappingDim: -1, workgroupMappingValue: -1)| GEMM(mac_m: 64, mac_n: 64, mac_k: 64, wave_m: 32, wave_n: 32, wave_k: 2, wave_b: 1, workgroup_size_x: 128, workgroup_size_y: 2, workgroupRemapXCC: False, workgroupRemapXCCValue: -1, unroll_x: 0, unroll_y: 0, load_A: BufferToLDSViaVGPR, load_B: BufferToLDSViaVGPR, storeLDS_D: True, betaInFma: True, scheduler: Priority, schedulerCost: LinearWeighted, prefetch: False, prefetchInFlight: 2, prefetchLDSFactor: 0, prefetchMixMemOps: False, loadScale_A: BufferToVGPR, loadScale_B: BufferToVGPR, swizzleScale: False, swizzleTileSize: {'m': 0, 'k': 0, 'n': 0, 'l': 0}, prefetchScale: False, streamK: True, numWGs: 304, streamKTwoTile: False, streamKTwoTileDPFirst: True, architecture: {'ArchString': 'gfx942', 'Xnack': False, 'Sramecc': True}, matchMemoryAccess: True, version: ) | GEMMProblem(M=3072, N=4096, K=4096, alpha=2, beta=0.5, types={'type_A': 'float', 'type_B': 'float', 'type_C': 'float', 'type_D': 'float', 'type_acc': 'float', 'trans_A': 'N', 'trans_B': 'T', 'scale_A': 'None', 'scaleType_A': 'None', 'scale_B': 'None', 'scaleType_B': 'None', 'scaleBlockSize': -1, 'scaleSkipPermlane': False, 'scaleShuffleTileA': [], 'scaleShuffleTileB': []}, scaleValue_A=1, scaleValue_B=1, initMode_A='DataInitMode(Bounded)', initMode_B='DataInitMode(Bounded)', initMode_C='DataInitMode(Bounded)', workgroupMappingDim=-1, workgroupMappingValue=-1)GEMMSolution(mac_m=64, mac_n=64, mac_k=64, wave_m=32, wave_n=32, wave_k=2, wave_b=1, workgroup_size_x=128, workgroup_size_y=2, workgroupRemapXCC=False, workgroupRemapXCCValue=-1, unroll_x=0, unroll_y=0, load_A='BufferToLDSViaVGPR', load_B='BufferToLDSViaVGPR', storeLDS_D=True, betaInFma=True, scheduler='Priority', schedulerCost='LinearWeighted', prefetch=False, prefetchInFlight=2, prefetchLDSFactor=0, prefetchMixMemOps=False, loadScale_A='BufferToVGPR', loadScale_B='BufferToVGPR', swizzleScale=False, swizzleTileSize={'m': 0, 'k': 0, 'n': 0, 'l': 0}, prefetchScale=False, streamK=True, numWGs=304, streamKTwoTile=False, streamKTwoTileDPFirst=True, architecture={'ArchString': 'gfx942', 'Xnack': False, 'Sramecc': True}, matchMemoryAccess=True, version='')
- SGPR: 72 -> 74 (+2)
	| 3. FloatsGEMM(M: 3072, N: 4096, K: 4096, alpha: 2, beta: 0.5, types: {'type_A': 'float', 'type_B': 'float', 'type_C': 'float', 'type_D': 'float', 'type_acc': 'float', 'trans_A': 'N', 'trans_B': 'T', 'scale_A': 'None', 'scaleType_A': 'None', 'scale_B': 'None', 'scaleType_B': 'None', 'scaleBlockSize': -1, 'scaleSkipPermlane': False, 'scaleShuffleTileA': [], 'scaleShuffleTileB': []}, scaleValue_A: 1, scaleValue_B: 1, initMode_A: DataInitMode(Bounded), initMode_B: DataInitMode(Bounded), initMode_C: DataInitMode(Bounded), workgroupMappingDim: -1, workgroupMappingValue: -1)| GEMM(mac_m: 64, mac_n: 64, mac_k: 64, wave_m: 32, wave_n: 32, wave_k: 2, wave_b: 1, workgroup_size_x: 128, workgroup_size_y: 2, workgroupRemapXCC: False, workgroupRemapXCCValue: -1, unroll_x: 0, unroll_y: 0, load_A: BufferToLDSViaVGPR, load_B: BufferToLDSViaVGPR, storeLDS_D: True, betaInFma: True, scheduler: Priority, schedulerCost: LinearWeighted, prefetch: False, prefetchInFlight: 2, prefetchLDSFactor: 0, prefetchMixMemOps: False, loadScale_A: BufferToVGPR, loadScale_B: BufferToVGPR, swizzleScale: False, swizzleTileSize: {'m': 0, 'k': 0, 'n': 0, 'l': 0}, prefetchScale: False, streamK: True, numWGs: 304, streamKTwoTile: True, streamKTwoTileDPFirst: False, architecture: {'ArchString': 'gfx942', 'Xnack': False, 'Sramecc': True}, matchMemoryAccess: True, version: ) | GEMMProblem(M=3072, N=4096, K=4096, alpha=2, beta=0.5, types={'type_A': 'float', 'type_B': 'float', 'type_C': 'float', 'type_D': 'float', 'type_acc': 'float', 'trans_A': 'N', 'trans_B': 'T', 'scale_A': 'None', 'scaleType_A': 'None', 'scale_B': 'None', 'scaleType_B': 'None', 'scaleBlockSize': -1, 'scaleSkipPermlane': False, 'scaleShuffleTileA': [], 'scaleShuffleTileB': []}, scaleValue_A=1, scaleValue_B=1, initMode_A='DataInitMode(Bounded)', initMode_B='DataInitMode(Bounded)', initMode_C='DataInitMode(Bounded)', workgroupMappingDim=-1, workgroupMappingValue=-1)GEMMSolution(mac_m=64, mac_n=64, mac_k=64, wave_m=32, wave_n=32, wave_k=2, wave_b=1, workgroup_size_x=128, workgroup_size_y=2, workgroupRemapXCC=False, workgroupRemapXCCValue=-1, unroll_x=0, unroll_y=0, load_A='BufferToLDSViaVGPR', load_B='BufferToLDSViaVGPR', storeLDS_D=True, betaInFma=True, scheduler='Priority', schedulerCost='LinearWeighted', prefetch=False, prefetchInFlight=2, prefetchLDSFactor=0, prefetchMixMemOps=False, loadScale_A='BufferToVGPR', loadScale_B='BufferToVGPR', swizzleScale=False, swizzleTileSize={'m': 0, 'k': 0, 'n': 0, 'l': 0}, prefetchScale=False, streamK=True, numWGs=304, streamKTwoTile=True, streamKTwoTileDPFirst=False, architecture={'ArchString': 'gfx942', 'Xnack': False, 'Sramecc': True}, matchMemoryAccess=True, version='')
- SGPR: 80 -> 82 (+2)
	| 3. FloatsGEMM(M: 7680, N: 8448, K: 8192, alpha: 2, beta: 0.5, types: {'type_A': 'half', 'type_B': 'half', 'type_C': 'half', 'type_D': 'half', 'type_acc': 'float', 'trans_A': 'N', 'trans_B': 'T', 'scale_A': 'None', 'scaleType_A': 'None', 'scale_B': 'None', 'scaleType_B': 'None', 'scaleBlockSize': -1, 'scaleSkipPermlane': False, 'scaleShuffleTileA': [], 'scaleShuffleTileB': []}, scaleValue_A: 1, scaleValue_B: 1, initMode_A: DataInitMode(Bounded), initMode_B: DataInitMode(Bounded), initMode_C: DataInitMode(Bounded), workgroupMappingDim: -1, workgroupMappingValue: -1)| GEMM(mac_m: 128, mac_n: 256, mac_k: 16, wave_m: 32, wave_n: 32, wave_k: 8, wave_b: 1, workgroup_size_x: 128, workgroup_size_y: 2, workgroupRemapXCC: False, workgroupRemapXCCValue: -1, unroll_x: 0, unroll_y: 0, load_A: BufferToLDSViaVGPR, load_B: BufferToLDSViaVGPR, storeLDS_D: True, betaInFma: True, scheduler: Priority, schedulerCost: LinearWeighted, prefetch: False, prefetchInFlight: 2, prefetchLDSFactor: 0, prefetchMixMemOps: False, loadScale_A: BufferToVGPR, loadScale_B: BufferToVGPR, swizzleScale: False, swizzleTileSize: {'m': 0, 'k': 0, 'n': 0, 'l': 0}, prefetchScale: False, streamK: True, numWGs: 304, streamKTwoTile: False, streamKTwoTileDPFirst: False, architecture: {'ArchString': 'gfx942', 'Xnack': False, 'Sramecc': True}, matchMemoryAccess: True, version: ) | GEMMProblem(M=7680, N=8448, K=8192, alpha=2, beta=0.5, types={'type_A': 'half', 'type_B': 'half', 'type_C': 'half', 'type_D': 'half', 'type_acc': 'float', 'trans_A': 'N', 'trans_B': 'T', 'scale_A': 'None', 'scaleType_A': 'None', 'scale_B': 'None', 'scaleType_B': 'None', 'scaleBlockSize': -1, 'scaleSkipPermlane': False, 'scaleShuffleTileA': [], 'scaleShuffleTileB': []}, scaleValue_A=1, scaleValue_B=1, initMode_A='DataInitMode(Bounded)', initMode_B='DataInitMode(Bounded)', initMode_C='DataInitMode(Bounded)', workgroupMappingDim=-1, workgroupMappingValue=-1)GEMMSolution(mac_m=128, mac_n=256, mac_k=16, wave_m=32, wave_n=32, wave_k=8, wave_b=1, workgroup_size_x=128, workgroup_size_y=2, workgroupRemapXCC=False, workgroupRemapXCCValue=-1, unroll_x=0, unroll_y=0, load_A='BufferToLDSViaVGPR', load_B='BufferToLDSViaVGPR', storeLDS_D=True, betaInFma=True, scheduler='Priority', schedulerCost='LinearWeighted', prefetch=False, prefetchInFlight=2, prefetchLDSFactor=0, prefetchMixMemOps=False, loadScale_A='BufferToVGPR', loadScale_B='BufferToVGPR', swizzleScale=False, swizzleTileSize={'m': 0, 'k': 0, 'n': 0, 'l': 0}, prefetchScale=False, streamK=True, numWGs=304, streamKTwoTile=False, streamKTwoTileDPFirst=False, architecture={'ArchString': 'gfx942', 'Xnack': False, 'Sramecc': True}, matchMemoryAccess=True, version='')
- SGPR: 84 -> 86 (+2)
	| 3. FloatsGEMM(M: 7680, N: 8448, K: 8192, alpha: 2, beta: 0.5, types: {'type_A': 'half', 'type_B': 'half', 'type_C': 'half', 'type_D': 'half', 'type_acc': 'float', 'trans_A': 'N', 'trans_B': 'T', 'scale_A': 'None', 'scaleType_A': 'None', 'scale_B': 'None', 'scaleType_B': 'None', 'scaleBlockSize': -1, 'scaleSkipPermlane': False, 'scaleShuffleTileA': [], 'scaleShuffleTileB': []}, scaleValue_A: 1, scaleValue_B: 1, initMode_A: DataInitMode(Bounded), initMode_B: DataInitMode(Bounded), initMode_C: DataInitMode(Bounded), workgroupMappingDim: -1, workgroupMappingValue: -1)| GEMM(mac_m: 128, mac_n: 256, mac_k: 16, wave_m: 32, wave_n: 32, wave_k: 8, wave_b: 1, workgroup_size_x: 128, workgroup_size_y: 2, workgroupRemapXCC: False, workgroupRemapXCCValue: -1, unroll_x: 0, unroll_y: 0, load_A: BufferToLDSViaVGPR, load_B: BufferToLDSViaVGPR, storeLDS_D: True, betaInFma: True, scheduler: Priority, schedulerCost: LinearWeighted, prefetch: False, prefetchInFlight: 2, prefetchLDSFactor: 0, prefetchMixMemOps: False, loadScale_A: BufferToVGPR, loadScale_B: BufferToVGPR, swizzleScale: False, swizzleTileSize: {'m': 0, 'k': 0, 'n': 0, 'l': 0}, prefetchScale: False, streamK: True, numWGs: 304, streamKTwoTile: False, streamKTwoTileDPFirst: True, architecture: {'ArchString': 'gfx942', 'Xnack': False, 'Sramecc': True}, matchMemoryAccess: True, version: ) | GEMMProblem(M=7680, N=8448, K=8192, alpha=2, beta=0.5, types={'type_A': 'half', 'type_B': 'half', 'type_C': 'half', 'type_D': 'half', 'type_acc': 'float', 'trans_A': 'N', 'trans_B': 'T', 'scale_A': 'None', 'scaleType_A': 'None', 'scale_B': 'None', 'scaleType_B': 'None', 'scaleBlockSize': -1, 'scaleSkipPermlane': False, 'scaleShuffleTileA': [], 'scaleShuffleTileB': []}, scaleValue_A=1, scaleValue_B=1, initMode_A='DataInitMode(Bounded)', initMode_B='DataInitMode(Bounded)', initMode_C='DataInitMode(Bounded)', workgroupMappingDim=-1, workgroupMappingValue=-1)GEMMSolution(mac_m=128, mac_n=256, mac_k=16, wave_m=32, wave_n=32, wave_k=8, wave_b=1, workgroup_size_x=128, workgroup_size_y=2, workgroupRemapXCC=False, workgroupRemapXCCValue=-1, unroll_x=0, unroll_y=0, load_A='BufferToLDSViaVGPR', load_B='BufferToLDSViaVGPR', storeLDS_D=True, betaInFma=True, scheduler='Priority', schedulerCost='LinearWeighted', prefetch=False, prefetchInFlight=2, prefetchLDSFactor=0, prefetchMixMemOps=False, loadScale_A='BufferToVGPR', loadScale_B='BufferToVGPR', swizzleScale=False, swizzleTileSize={'m': 0, 'k': 0, 'n': 0, 'l': 0}, prefetchScale=False, streamK=True, numWGs=304, streamKTwoTile=False, streamKTwoTileDPFirst=True, architecture={'ArchString': 'gfx942', 'Xnack': False, 'Sramecc': True}, matchMemoryAccess=True, version='')
- SGPR: 84 -> 86 (+2)
	| 3. FloatsGEMM(M: 7680, N: 8448, K: 8192, alpha: 2, beta: 0.5, types: {'type_A': 'half', 'type_B': 'half', 'type_C': 'half', 'type_D': 'half', 'type_acc': 'float', 'trans_A': 'N', 'trans_B': 'T', 'scale_A': 'None', 'scaleType_A': 'None', 'scale_B': 'None', 'scaleType_B': 'None', 'scaleBlockSize': -1, 'scaleSkipPermlane': False, 'scaleShuffleTileA': [], 'scaleShuffleTileB': []}, scaleValue_A: 1, scaleValue_B: 1, initMode_A: DataInitMode(Bounded), initMode_B: DataInitMode(Bounded), initMode_C: DataInitMode(Bounded), workgroupMappingDim: -1, workgroupMappingValue: -1)| GEMM(mac_m: 128, mac_n: 256, mac_k: 16, wave_m: 32, wave_n: 32, wave_k: 8, wave_b: 1, workgroup_size_x: 128, workgroup_size_y: 2, workgroupRemapXCC: False, workgroupRemapXCCValue: -1, unroll_x: 0, unroll_y: 0, load_A: BufferToLDSViaVGPR, load_B: BufferToLDSViaVGPR, storeLDS_D: True, betaInFma: True, scheduler: Priority, schedulerCost: LinearWeighted, prefetch: False, prefetchInFlight: 2, prefetchLDSFactor: 0, prefetchMixMemOps: False, loadScale_A: BufferToVGPR, loadScale_B: BufferToVGPR, swizzleScale: False, swizzleTileSize: {'m': 0, 'k': 0, 'n': 0, 'l': 0}, prefetchScale: False, streamK: True, numWGs: 304, streamKTwoTile: True, streamKTwoTileDPFirst: False, architecture: {'ArchString': 'gfx942', 'Xnack': False, 'Sramecc': True}, matchMemoryAccess: True, version: ) | GEMMProblem(M=7680, N=8448, K=8192, alpha=2, beta=0.5, types={'type_A': 'half', 'type_B': 'half', 'type_C': 'half', 'type_D': 'half', 'type_acc': 'float', 'trans_A': 'N', 'trans_B': 'T', 'scale_A': 'None', 'scaleType_A': 'None', 'scale_B': 'None', 'scaleType_B': 'None', 'scaleBlockSize': -1, 'scaleSkipPermlane': False, 'scaleShuffleTileA': [], 'scaleShuffleTileB': []}, scaleValue_A=1, scaleValue_B=1, initMode_A='DataInitMode(Bounded)', initMode_B='DataInitMode(Bounded)', initMode_C='DataInitMode(Bounded)', workgroupMappingDim=-1, workgroupMappingValue=-1)GEMMSolution(mac_m=128, mac_n=256, mac_k=16, wave_m=32, wave_n=32, wave_k=8, wave_b=1, workgroup_size_x=128, workgroup_size_y=2, workgroupRemapXCC=False, workgroupRemapXCCValue=-1, unroll_x=0, unroll_y=0, load_A='BufferToLDSViaVGPR', load_B='BufferToLDSViaVGPR', storeLDS_D=True, betaInFma=True, scheduler='Priority', schedulerCost='LinearWeighted', prefetch=False, prefetchInFlight=2, prefetchLDSFactor=0, prefetchMixMemOps=False, loadScale_A='BufferToVGPR', loadScale_B='BufferToVGPR', swizzleScale=False, swizzleTileSize={'m': 0, 'k': 0, 'n': 0, 'l': 0}, prefetchScale=False, streamK=True, numWGs=304, streamKTwoTile=True, streamKTwoTileDPFirst=False, architecture={'ArchString': 'gfx942', 'Xnack': False, 'Sramecc': True}, matchMemoryAccess=True, version='')
- SGPR: 80 -> 82 (+2)
	| 3. FloatsGEMM(M: 7680, N: 8448, K: 8448, alpha: 2, beta: 0.5, types: {'type_A': 'half', 'type_B': 'half', 'type_C': 'half', 'type_D': 'half', 'type_acc': 'float', 'trans_A': 'N', 'trans_B': 'T', 'scale_A': 'None', 'scaleType_A': 'None', 'scale_B': 'None', 'scaleType_B': 'None', 'scaleBlockSize': -1, 'scaleSkipPermlane': False, 'scaleShuffleTileA': [], 'scaleShuffleTileB': []}, scaleValue_A: 1, scaleValue_B: 1, initMode_A: DataInitMode(Bounded), initMode_B: DataInitMode(Bounded), initMode_C: DataInitMode(Bounded), workgroupMappingDim: -1, workgroupMappingValue: -1)| GEMM(mac_m: 128, mac_n: 128, mac_k: 16, wave_m: 32, wave_n: 32, wave_k: 8, wave_b: 1, workgroup_size_x: 128, workgroup_size_y: 2, workgroupRemapXCC: False, workgroupRemapXCCValue: -1, unroll_x: 0, unroll_y: 0, load_A: BufferToLDSViaVGPR, load_B: BufferToLDSViaVGPR, storeLDS_D: True, betaInFma: True, scheduler: Priority, schedulerCost: LinearWeighted, prefetch: False, prefetchInFlight: 0, prefetchLDSFactor: 0, prefetchMixMemOps: False, loadScale_A: BufferToVGPR, loadScale_B: BufferToVGPR, swizzleScale: False, swizzleTileSize: {'m': 0, 'k': 0, 'n': 0, 'l': 0}, prefetchScale: False, streamK: True, numWGs: 304, streamKTwoTile: False, streamKTwoTileDPFirst: False, architecture: {'ArchString': 'gfx942', 'Xnack': False, 'Sramecc': True}, matchMemoryAccess: True, version: ) | GEMMProblem(M=7680, N=8448, K=8448, alpha=2, beta=0.5, types={'type_A': 'half', 'type_B': 'half', 'type_C': 'half', 'type_D': 'half', 'type_acc': 'float', 'trans_A': 'N', 'trans_B': 'T', 'scale_A': 'None', 'scaleType_A': 'None', 'scale_B': 'None', 'scaleType_B': 'None', 'scaleBlockSize': -1, 'scaleSkipPermlane': False, 'scaleShuffleTileA': [], 'scaleShuffleTileB': []}, scaleValue_A=1, scaleValue_B=1, initMode_A='DataInitMode(Bounded)', initMode_B='DataInitMode(Bounded)', initMode_C='DataInitMode(Bounded)', workgroupMappingDim=-1, workgroupMappingValue=-1)GEMMSolution(mac_m=128, mac_n=128, mac_k=16, wave_m=32, wave_n=32, wave_k=8, wave_b=1, workgroup_size_x=128, workgroup_size_y=2, workgroupRemapXCC=False, workgroupRemapXCCValue=-1, unroll_x=0, unroll_y=0, load_A='BufferToLDSViaVGPR', load_B='BufferToLDSViaVGPR', storeLDS_D=True, betaInFma=True, scheduler='Priority', schedulerCost='LinearWeighted', prefetch=False, prefetchInFlight=0, prefetchLDSFactor=0, prefetchMixMemOps=False, loadScale_A='BufferToVGPR', loadScale_B='BufferToVGPR', swizzleScale=False, swizzleTileSize={'m': 0, 'k': 0, 'n': 0, 'l': 0}, prefetchScale=False, streamK=True, numWGs=304, streamKTwoTile=False, streamKTwoTileDPFirst=False, architecture={'ArchString': 'gfx942', 'Xnack': False, 'Sramecc': True}, matchMemoryAccess=True, version='')
- SGPR: 84 -> 86 (+2)
	| 3. FloatsGEMM(M: 7680, N: 8448, K: 8448, alpha: 2, beta: 0.5, types: {'type_A': 'half', 'type_B': 'half', 'type_C': 'half', 'type_D': 'half', 'type_acc': 'float', 'trans_A': 'N', 'trans_B': 'T', 'scale_A': 'None', 'scaleType_A': 'None', 'scale_B': 'None', 'scaleType_B': 'None', 'scaleBlockSize': -1, 'scaleSkipPermlane': False, 'scaleShuffleTileA': [], 'scaleShuffleTileB': []}, scaleValue_A: 1, scaleValue_B: 1, initMode_A: DataInitMode(Bounded), initMode_B: DataInitMode(Bounded), initMode_C: DataInitMode(Bounded), workgroupMappingDim: -1, workgroupMappingValue: -1)| GEMM(mac_m: 128, mac_n: 128, mac_k: 16, wave_m: 32, wave_n: 32, wave_k: 8, wave_b: 1, workgroup_size_x: 128, workgroup_size_y: 2, workgroupRemapXCC: False, workgroupRemapXCCValue: -1, unroll_x: 0, unroll_y: 0, load_A: BufferToLDSViaVGPR, load_B: BufferToLDSViaVGPR, storeLDS_D: True, betaInFma: True, scheduler: Priority, schedulerCost: LinearWeighted, prefetch: False, prefetchInFlight: 0, prefetchLDSFactor: 0, prefetchMixMemOps: False, loadScale_A: BufferToVGPR, loadScale_B: BufferToVGPR, swizzleScale: False, swizzleTileSize: {'m': 0, 'k': 0, 'n': 0, 'l': 0}, prefetchScale: False, streamK: True, numWGs: 304, streamKTwoTile: False, streamKTwoTileDPFirst: True, architecture: {'ArchString': 'gfx942', 'Xnack': False, 'Sramecc': True}, matchMemoryAccess: True, version: ) | GEMMProblem(M=7680, N=8448, K=8448, alpha=2, beta=0.5, types={'type_A': 'half', 'type_B': 'half', 'type_C': 'half', 'type_D': 'half', 'type_acc': 'float', 'trans_A': 'N', 'trans_B': 'T', 'scale_A': 'None', 'scaleType_A': 'None', 'scale_B': 'None', 'scaleType_B': 'None', 'scaleBlockSize': -1, 'scaleSkipPermlane': False, 'scaleShuffleTileA': [], 'scaleShuffleTileB': []}, scaleValue_A=1, scaleValue_B=1, initMode_A='DataInitMode(Bounded)', initMode_B='DataInitMode(Bounded)', initMode_C='DataInitMode(Bounded)', workgroupMappingDim=-1, workgroupMappingValue=-1)GEMMSolution(mac_m=128, mac_n=128, mac_k=16, wave_m=32, wave_n=32, wave_k=8, wave_b=1, workgroup_size_x=128, workgroup_size_y=2, workgroupRemapXCC=False, workgroupRemapXCCValue=-1, unroll_x=0, unroll_y=0, load_A='BufferToLDSViaVGPR', load_B='BufferToLDSViaVGPR', storeLDS_D=True, betaInFma=True, scheduler='Priority', schedulerCost='LinearWeighted', prefetch=False, prefetchInFlight=0, prefetchLDSFactor=0, prefetchMixMemOps=False, loadScale_A='BufferToVGPR', loadScale_B='BufferToVGPR', swizzleScale=False, swizzleTileSize={'m': 0, 'k': 0, 'n': 0, 'l': 0}, prefetchScale=False, streamK=True, numWGs=304, streamKTwoTile=False, streamKTwoTileDPFirst=True, architecture={'ArchString': 'gfx942', 'Xnack': False, 'Sramecc': True}, matchMemoryAccess=True, version='')
- SGPR: 84 -> 86 (+2)
	| 3. FloatsGEMM(M: 7680, N: 8448, K: 8448, alpha: 2, beta: 0.5, types: {'type_A': 'half', 'type_B': 'half', 'type_C': 'half', 'type_D': 'half', 'type_acc': 'float', 'trans_A': 'N', 'trans_B': 'T', 'scale_A': 'None', 'scaleType_A': 'None', 'scale_B': 'None', 'scaleType_B': 'None', 'scaleBlockSize': -1, 'scaleSkipPermlane': False, 'scaleShuffleTileA': [], 'scaleShuffleTileB': []}, scaleValue_A: 1, scaleValue_B: 1, initMode_A: DataInitMode(Bounded), initMode_B: DataInitMode(Bounded), initMode_C: DataInitMode(Bounded), workgroupMappingDim: -1, workgroupMappingValue: -1)| GEMM(mac_m: 128, mac_n: 128, mac_k: 16, wave_m: 32, wave_n: 32, wave_k: 8, wave_b: 1, workgroup_size_x: 128, workgroup_size_y: 2, workgroupRemapXCC: False, workgroupRemapXCCValue: -1, unroll_x: 0, unroll_y: 0, load_A: BufferToLDSViaVGPR, load_B: BufferToLDSViaVGPR, storeLDS_D: True, betaInFma: True, scheduler: Priority, schedulerCost: LinearWeighted, prefetch: False, prefetchInFlight: 0, prefetchLDSFactor: 0, prefetchMixMemOps: False, loadScale_A: BufferToVGPR, loadScale_B: BufferToVGPR, swizzleScale: False, swizzleTileSize: {'m': 0, 'k': 0, 'n': 0, 'l': 0}, prefetchScale: False, streamK: True, numWGs: 304, streamKTwoTile: True, streamKTwoTileDPFirst: False, architecture: {'ArchString': 'gfx942', 'Xnack': False, 'Sramecc': True}, matchMemoryAccess: True, version: ) | GEMMProblem(M=7680, N=8448, K=8448, alpha=2, beta=0.5, types={'type_A': 'half', 'type_B': 'half', 'type_C': 'half', 'type_D': 'half', 'type_acc': 'float', 'trans_A': 'N', 'trans_B': 'T', 'scale_A': 'None', 'scaleType_A': 'None', 'scale_B': 'None', 'scaleType_B': 'None', 'scaleBlockSize': -1, 'scaleSkipPermlane': False, 'scaleShuffleTileA': [], 'scaleShuffleTileB': []}, scaleValue_A=1, scaleValue_B=1, initMode_A='DataInitMode(Bounded)', initMode_B='DataInitMode(Bounded)', initMode_C='DataInitMode(Bounded)', workgroupMappingDim=-1, workgroupMappingValue=-1)GEMMSolution(mac_m=128, mac_n=128, mac_k=16, wave_m=32, wave_n=32, wave_k=8, wave_b=1, workgroup_size_x=128, workgroup_size_y=2, workgroupRemapXCC=False, workgroupRemapXCCValue=-1, unroll_x=0, unroll_y=0, load_A='BufferToLDSViaVGPR', load_B='BufferToLDSViaVGPR', storeLDS_D=True, betaInFma=True, scheduler='Priority', schedulerCost='LinearWeighted', prefetch=False, prefetchInFlight=0, prefetchLDSFactor=0, prefetchMixMemOps=False, loadScale_A='BufferToVGPR', loadScale_B='BufferToVGPR', swizzleScale=False, swizzleTileSize={'m': 0, 'k': 0, 'n': 0, 'l': 0}, prefetchScale=False, streamK=True, numWGs=304, streamKTwoTile=True, streamKTwoTileDPFirst=False, architecture={'ArchString': 'gfx942', 'Xnack': False, 'Sramecc': True}, matchMemoryAccess=True, version='')
- SGPR: 80 -> 82 (+2)
	| 3. FloatsGEMM(M: 7680, N: 8448, K: 8448, alpha: 2, beta: 0.5, types: {'type_A': 'half', 'type_B': 'half', 'type_C': 'half', 'type_D': 'half', 'type_acc': 'float', 'trans_A': 'N', 'trans_B': 'T', 'scale_A': 'None', 'scaleType_A': 'None', 'scale_B': 'None', 'scaleType_B': 'None', 'scaleBlockSize': -1, 'scaleSkipPermlane': False, 'scaleShuffleTileA': [], 'scaleShuffleTileB': []}, scaleValue_A: 1, scaleValue_B: 1, initMode_A: DataInitMode(Bounded), initMode_B: DataInitMode(Bounded), initMode_C: DataInitMode(Bounded), workgroupMappingDim: -1, workgroupMappingValue: -1)| GEMM(mac_m: 128, mac_n: 128, mac_k: 16, wave_m: 32, wave_n: 32, wave_k: 8, wave_b: 1, workgroup_size_x: 128, workgroup_size_y: 2, workgroupRemapXCC: False, workgroupRemapXCCValue: -1, unroll_x: 0, unroll_y: 0, load_A: BufferToLDSViaVGPR, load_B: BufferToLDSViaVGPR, storeLDS_D: True, betaInFma: True, scheduler: Priority, schedulerCost: LinearWeighted, prefetch: True, prefetchInFlight: 2, prefetchLDSFactor: 2, prefetchMixMemOps: False, loadScale_A: BufferToVGPR, loadScale_B: BufferToVGPR, swizzleScale: False, swizzleTileSize: {'m': 0, 'k': 0, 'n': 0, 'l': 0}, prefetchScale: False, streamK: True, numWGs: 304, streamKTwoTile: False, streamKTwoTileDPFirst: False, architecture: {'ArchString': 'gfx942', 'Xnack': False, 'Sramecc': True}, matchMemoryAccess: True, version: ) | GEMMProblem(M=7680, N=8448, K=8448, alpha=2, beta=0.5, types={'type_A': 'half', 'type_B': 'half', 'type_C': 'half', 'type_D': 'half', 'type_acc': 'float', 'trans_A': 'N', 'trans_B': 'T', 'scale_A': 'None', 'scaleType_A': 'None', 'scale_B': 'None', 'scaleType_B': 'None', 'scaleBlockSize': -1, 'scaleSkipPermlane': False, 'scaleShuffleTileA': [], 'scaleShuffleTileB': []}, scaleValue_A=1, scaleValue_B=1, initMode_A='DataInitMode(Bounded)', initMode_B='DataInitMode(Bounded)', initMode_C='DataInitMode(Bounded)', workgroupMappingDim=-1, workgroupMappingValue=-1)GEMMSolution(mac_m=128, mac_n=128, mac_k=16, wave_m=32, wave_n=32, wave_k=8, wave_b=1, workgroup_size_x=128, workgroup_size_y=2, workgroupRemapXCC=False, workgroupRemapXCCValue=-1, unroll_x=0, unroll_y=0, load_A='BufferToLDSViaVGPR', load_B='BufferToLDSViaVGPR', storeLDS_D=True, betaInFma=True, scheduler='Priority', schedulerCost='LinearWeighted', prefetch=True, prefetchInFlight=2, prefetchLDSFactor=2, prefetchMixMemOps=False, loadScale_A='BufferToVGPR', loadScale_B='BufferToVGPR', swizzleScale=False, swizzleTileSize={'m': 0, 'k': 0, 'n': 0, 'l': 0}, prefetchScale=False, streamK=True, numWGs=304, streamKTwoTile=False, streamKTwoTileDPFirst=False, architecture={'ArchString': 'gfx942', 'Xnack': False, 'Sramecc': True}, matchMemoryAccess=True, version='')
- SGPR: 80 -> 82 (+2)
	| 3. FloatsGEMM(M: 7680, N: 8448, K: 8448, alpha: 2, beta: 0.5, types: {'type_A': 'half', 'type_B': 'half', 'type_C': 'half', 'type_D': 'half', 'type_acc': 'float', 'trans_A': 'N', 'trans_B': 'T', 'scale_A': 'None', 'scaleType_A': 'None', 'scale_B': 'None', 'scaleType_B': 'None', 'scaleBlockSize': -1, 'scaleSkipPermlane': False, 'scaleShuffleTileA': [], 'scaleShuffleTileB': []}, scaleValue_A: 1, scaleValue_B: 1, initMode_A: DataInitMode(Bounded), initMode_B: DataInitMode(Bounded), initMode_C: DataInitMode(Bounded), workgroupMappingDim: -1, workgroupMappingValue: -1)| GEMM(mac_m: 128, mac_n: 128, mac_k: 32, wave_m: 32, wave_n: 32, wave_k: 8, wave_b: 1, workgroup_size_x: 128, workgroup_size_y: 2, workgroupRemapXCC: False, workgroupRemapXCCValue: -1, unroll_x: 0, unroll_y: 0, load_A: BufferToLDSViaVGPR, load_B: BufferToLDSViaVGPR, storeLDS_D: True, betaInFma: True, scheduler: Priority, schedulerCost: LinearWeighted, prefetch: False, prefetchInFlight: 0, prefetchLDSFactor: 0, prefetchMixMemOps: False, loadScale_A: BufferToVGPR, loadScale_B: BufferToVGPR, swizzleScale: False, swizzleTileSize: {'m': 0, 'k': 0, 'n': 0, 'l': 0}, prefetchScale: False, streamK: True, numWGs: 304, streamKTwoTile: False, streamKTwoTileDPFirst: False, architecture: {'ArchString': 'gfx942', 'Xnack': False, 'Sramecc': True}, matchMemoryAccess: True, version: ) | GEMMProblem(M=7680, N=8448, K=8448, alpha=2, beta=0.5, types={'type_A': 'half', 'type_B': 'half', 'type_C': 'half', 'type_D': 'half', 'type_acc': 'float', 'trans_A': 'N', 'trans_B': 'T', 'scale_A': 'None', 'scaleType_A': 'None', 'scale_B': 'None', 'scaleType_B': 'None', 'scaleBlockSize': -1, 'scaleSkipPermlane': False, 'scaleShuffleTileA': [], 'scaleShuffleTileB': []}, scaleValue_A=1, scaleValue_B=1, initMode_A='DataInitMode(Bounded)', initMode_B='DataInitMode(Bounded)', initMode_C='DataInitMode(Bounded)', workgroupMappingDim=-1, workgroupMappingValue=-1)GEMMSolution(mac_m=128, mac_n=128, mac_k=32, wave_m=32, wave_n=32, wave_k=8, wave_b=1, workgroup_size_x=128, workgroup_size_y=2, workgroupRemapXCC=False, workgroupRemapXCCValue=-1, unroll_x=0, unroll_y=0, load_A='BufferToLDSViaVGPR', load_B='BufferToLDSViaVGPR', storeLDS_D=True, betaInFma=True, scheduler='Priority', schedulerCost='LinearWeighted', prefetch=False, prefetchInFlight=0, prefetchLDSFactor=0, prefetchMixMemOps=False, loadScale_A='BufferToVGPR', loadScale_B='BufferToVGPR', swizzleScale=False, swizzleTileSize={'m': 0, 'k': 0, 'n': 0, 'l': 0}, prefetchScale=False, streamK=True, numWGs=304, streamKTwoTile=False, streamKTwoTileDPFirst=False, architecture={'ArchString': 'gfx942', 'Xnack': False, 'Sramecc': True}, matchMemoryAccess=True, version='')
- SGPR: 84 -> 86 (+2)
	| 3. FloatsGEMM(M: 7680, N: 8448, K: 8448, alpha: 2, beta: 0.5, types: {'type_A': 'half', 'type_B': 'half', 'type_C': 'half', 'type_D': 'half', 'type_acc': 'float', 'trans_A': 'N', 'trans_B': 'T', 'scale_A': 'None', 'scaleType_A': 'None', 'scale_B': 'None', 'scaleType_B': 'None', 'scaleBlockSize': -1, 'scaleSkipPermlane': False, 'scaleShuffleTileA': [], 'scaleShuffleTileB': []}, scaleValue_A: 1, scaleValue_B: 1, initMode_A: DataInitMode(Bounded), initMode_B: DataInitMode(Bounded), initMode_C: DataInitMode(Bounded), workgroupMappingDim: -1, workgroupMappingValue: -1)| GEMM(mac_m: 128, mac_n: 128, mac_k: 32, wave_m: 32, wave_n: 32, wave_k: 8, wave_b: 1, workgroup_size_x: 128, workgroup_size_y: 2, workgroupRemapXCC: False, workgroupRemapXCCValue: -1, unroll_x: 0, unroll_y: 0, load_A: BufferToLDSViaVGPR, load_B: BufferToLDSViaVGPR, storeLDS_D: True, betaInFma: True, scheduler: Priority, schedulerCost: LinearWeighted, prefetch: False, prefetchInFlight: 0, prefetchLDSFactor: 0, prefetchMixMemOps: False, loadScale_A: BufferToVGPR, loadScale_B: BufferToVGPR, swizzleScale: False, swizzleTileSize: {'m': 0, 'k': 0, 'n': 0, 'l': 0}, prefetchScale: False, streamK: True, numWGs: 304, streamKTwoTile: False, streamKTwoTileDPFirst: True, architecture: {'ArchString': 'gfx942', 'Xnack': False, 'Sramecc': True}, matchMemoryAccess: True, version: ) | GEMMProblem(M=7680, N=8448, K=8448, alpha=2, beta=0.5, types={'type_A': 'half', 'type_B': 'half', 'type_C': 'half', 'type_D': 'half', 'type_acc': 'float', 'trans_A': 'N', 'trans_B': 'T', 'scale_A': 'None', 'scaleType_A': 'None', 'scale_B': 'None', 'scaleType_B': 'None', 'scaleBlockSize': -1, 'scaleSkipPermlane': False, 'scaleShuffleTileA': [], 'scaleShuffleTileB': []}, scaleValue_A=1, scaleValue_B=1, initMode_A='DataInitMode(Bounded)', initMode_B='DataInitMode(Bounded)', initMode_C='DataInitMode(Bounded)', workgroupMappingDim=-1, workgroupMappingValue=-1)GEMMSolution(mac_m=128, mac_n=128, mac_k=32, wave_m=32, wave_n=32, wave_k=8, wave_b=1, workgroup_size_x=128, workgroup_size_y=2, workgroupRemapXCC=False, workgroupRemapXCCValue=-1, unroll_x=0, unroll_y=0, load_A='BufferToLDSViaVGPR', load_B='BufferToLDSViaVGPR', storeLDS_D=True, betaInFma=True, scheduler='Priority', schedulerCost='LinearWeighted', prefetch=False, prefetchInFlight=0, prefetchLDSFactor=0, prefetchMixMemOps=False, loadScale_A='BufferToVGPR', loadScale_B='BufferToVGPR', swizzleScale=False, swizzleTileSize={'m': 0, 'k': 0, 'n': 0, 'l': 0}, prefetchScale=False, streamK=True, numWGs=304, streamKTwoTile=False, streamKTwoTileDPFirst=True, architecture={'ArchString': 'gfx942', 'Xnack': False, 'Sramecc': True}, matchMemoryAccess=True, version='')
- SGPR: 84 -> 86 (+2)
	| 3. FloatsGEMM(M: 7680, N: 8448, K: 8448, alpha: 2, beta: 0.5, types: {'type_A': 'half', 'type_B': 'half', 'type_C': 'half', 'type_D': 'half', 'type_acc': 'float', 'trans_A': 'N', 'trans_B': 'T', 'scale_A': 'None', 'scaleType_A': 'None', 'scale_B': 'None', 'scaleType_B': 'None', 'scaleBlockSize': -1, 'scaleSkipPermlane': False, 'scaleShuffleTileA': [], 'scaleShuffleTileB': []}, scaleValue_A: 1, scaleValue_B: 1, initMode_A: DataInitMode(Bounded), initMode_B: DataInitMode(Bounded), initMode_C: DataInitMode(Bounded), workgroupMappingDim: -1, workgroupMappingValue: -1)| GEMM(mac_m: 128, mac_n: 128, mac_k: 32, wave_m: 32, wave_n: 32, wave_k: 8, wave_b: 1, workgroup_size_x: 128, workgroup_size_y: 2, workgroupRemapXCC: False, workgroupRemapXCCValue: -1, unroll_x: 0, unroll_y: 0, load_A: BufferToLDSViaVGPR, load_B: BufferToLDSViaVGPR, storeLDS_D: True, betaInFma: True, scheduler: Priority, schedulerCost: LinearWeighted, prefetch: False, prefetchInFlight: 0, prefetchLDSFactor: 0, prefetchMixMemOps: False, loadScale_A: BufferToVGPR, loadScale_B: BufferToVGPR, swizzleScale: False, swizzleTileSize: {'m': 0, 'k': 0, 'n': 0, 'l': 0}, prefetchScale: False, streamK: True, numWGs: 304, streamKTwoTile: True, streamKTwoTileDPFirst: False, architecture: {'ArchString': 'gfx942', 'Xnack': False, 'Sramecc': True}, matchMemoryAccess: True, version: ) | GEMMProblem(M=7680, N=8448, K=8448, alpha=2, beta=0.5, types={'type_A': 'half', 'type_B': 'half', 'type_C': 'half', 'type_D': 'half', 'type_acc': 'float', 'trans_A': 'N', 'trans_B': 'T', 'scale_A': 'None', 'scaleType_A': 'None', 'scale_B': 'None', 'scaleType_B': 'None', 'scaleBlockSize': -1, 'scaleSkipPermlane': False, 'scaleShuffleTileA': [], 'scaleShuffleTileB': []}, scaleValue_A=1, scaleValue_B=1, initMode_A='DataInitMode(Bounded)', initMode_B='DataInitMode(Bounded)', initMode_C='DataInitMode(Bounded)', workgroupMappingDim=-1, workgroupMappingValue=-1)GEMMSolution(mac_m=128, mac_n=128, mac_k=32, wave_m=32, wave_n=32, wave_k=8, wave_b=1, workgroup_size_x=128, workgroup_size_y=2, workgroupRemapXCC=False, workgroupRemapXCCValue=-1, unroll_x=0, unroll_y=0, load_A='BufferToLDSViaVGPR', load_B='BufferToLDSViaVGPR', storeLDS_D=True, betaInFma=True, scheduler='Priority', schedulerCost='LinearWeighted', prefetch=False, prefetchInFlight=0, prefetchLDSFactor=0, prefetchMixMemOps=False, loadScale_A='BufferToVGPR', loadScale_B='BufferToVGPR', swizzleScale=False, swizzleTileSize={'m': 0, 'k': 0, 'n': 0, 'l': 0}, prefetchScale=False, streamK=True, numWGs=304, streamKTwoTile=True, streamKTwoTileDPFirst=False, architecture={'ArchString': 'gfx942', 'Xnack': False, 'Sramecc': True}, matchMemoryAccess=True, version='')
- SGPR: 80 -> 82 (+2)
	| 3. FloatsGEMM(M: 7680, N: 8448, K: 8448, alpha: 2, beta: 0.5, types: {'type_A': 'half', 'type_B': 'half', 'type_C': 'half', 'type_D': 'half', 'type_acc': 'float', 'trans_A': 'N', 'trans_B': 'T', 'scale_A': 'None', 'scaleType_A': 'None', 'scale_B': 'None', 'scaleType_B': 'None', 'scaleBlockSize': -1, 'scaleSkipPermlane': False, 'scaleShuffleTileA': [], 'scaleShuffleTileB': []}, scaleValue_A: 1, scaleValue_B: 1, initMode_A: DataInitMode(Bounded), initMode_B: DataInitMode(Bounded), initMode_C: DataInitMode(Bounded), workgroupMappingDim: -1, workgroupMappingValue: -1)| GEMM(mac_m: 128, mac_n: 128, mac_k: 32, wave_m: 32, wave_n: 32, wave_k: 8, wave_b: 1, workgroup_size_x: 128, workgroup_size_y: 2, workgroupRemapXCC: False, workgroupRemapXCCValue: -1, unroll_x: 0, unroll_y: 0, load_A: BufferToLDSViaVGPR, load_B: BufferToLDSViaVGPR, storeLDS_D: True, betaInFma: True, scheduler: Priority, schedulerCost: LinearWeighted, prefetch: True, prefetchInFlight: 2, prefetchLDSFactor: 2, prefetchMixMemOps: False, loadScale_A: BufferToVGPR, loadScale_B: BufferToVGPR, swizzleScale: False, swizzleTileSize: {'m': 0, 'k': 0, 'n': 0, 'l': 0}, prefetchScale: False, streamK: True, numWGs: 304, streamKTwoTile: False, streamKTwoTileDPFirst: False, architecture: {'ArchString': 'gfx942', 'Xnack': False, 'Sramecc': True}, matchMemoryAccess: True, version: ) | GEMMProblem(M=7680, N=8448, K=8448, alpha=2, beta=0.5, types={'type_A': 'half', 'type_B': 'half', 'type_C': 'half', 'type_D': 'half', 'type_acc': 'float', 'trans_A': 'N', 'trans_B': 'T', 'scale_A': 'None', 'scaleType_A': 'None', 'scale_B': 'None', 'scaleType_B': 'None', 'scaleBlockSize': -1, 'scaleSkipPermlane': False, 'scaleShuffleTileA': [], 'scaleShuffleTileB': []}, scaleValue_A=1, scaleValue_B=1, initMode_A='DataInitMode(Bounded)', initMode_B='DataInitMode(Bounded)', initMode_C='DataInitMode(Bounded)', workgroupMappingDim=-1, workgroupMappingValue=-1)GEMMSolution(mac_m=128, mac_n=128, mac_k=32, wave_m=32, wave_n=32, wave_k=8, wave_b=1, workgroup_size_x=128, workgroup_size_y=2, workgroupRemapXCC=False, workgroupRemapXCCValue=-1, unroll_x=0, unroll_y=0, load_A='BufferToLDSViaVGPR', load_B='BufferToLDSViaVGPR', storeLDS_D=True, betaInFma=True, scheduler='Priority', schedulerCost='LinearWeighted', prefetch=True, prefetchInFlight=2, prefetchLDSFactor=2, prefetchMixMemOps=False, loadScale_A='BufferToVGPR', loadScale_B='BufferToVGPR', swizzleScale=False, swizzleTileSize={'m': 0, 'k': 0, 'n': 0, 'l': 0}, prefetchScale=False, streamK=True, numWGs=304, streamKTwoTile=False, streamKTwoTileDPFirst=False, architecture={'ArchString': 'gfx942', 'Xnack': False, 'Sramecc': True}, matchMemoryAccess=True, version='')
- SGPR: 80 -> 82 (+2)
	| 3. FloatsGEMM(M: 7680, N: 8448, K: 8448, alpha: 2, beta: 0.5, types: {'type_A': 'half', 'type_B': 'half', 'type_C': 'half', 'type_D': 'half', 'type_acc': 'float', 'trans_A': 'N', 'trans_B': 'T', 'scale_A': 'None', 'scaleType_A': 'None', 'scale_B': 'None', 'scaleType_B': 'None', 'scaleBlockSize': -1, 'scaleSkipPermlane': False, 'scaleShuffleTileA': [], 'scaleShuffleTileB': []}, scaleValue_A: 1, scaleValue_B: 1, initMode_A: DataInitMode(Bounded), initMode_B: DataInitMode(Bounded), initMode_C: DataInitMode(Bounded), workgroupMappingDim: -1, workgroupMappingValue: -1)| GEMM(mac_m: 128, mac_n: 128, mac_k: 64, wave_m: 32, wave_n: 32, wave_k: 8, wave_b: 1, workgroup_size_x: 128, workgroup_size_y: 2, workgroupRemapXCC: False, workgroupRemapXCCValue: -1, unroll_x: 0, unroll_y: 0, load_A: BufferToLDSViaVGPR, load_B: BufferToLDSViaVGPR, storeLDS_D: True, betaInFma: True, scheduler: Priority, schedulerCost: LinearWeighted, prefetch: False, prefetchInFlight: 0, prefetchLDSFactor: 0, prefetchMixMemOps: False, loadScale_A: BufferToVGPR, loadScale_B: BufferToVGPR, swizzleScale: False, swizzleTileSize: {'m': 0, 'k': 0, 'n': 0, 'l': 0}, prefetchScale: False, streamK: True, numWGs: 304, streamKTwoTile: False, streamKTwoTileDPFirst: False, architecture: {'ArchString': 'gfx942', 'Xnack': False, 'Sramecc': True}, matchMemoryAccess: True, version: ) | GEMMProblem(M=7680, N=8448, K=8448, alpha=2, beta=0.5, types={'type_A': 'half', 'type_B': 'half', 'type_C': 'half', 'type_D': 'half', 'type_acc': 'float', 'trans_A': 'N', 'trans_B': 'T', 'scale_A': 'None', 'scaleType_A': 'None', 'scale_B': 'None', 'scaleType_B': 'None', 'scaleBlockSize': -1, 'scaleSkipPermlane': False, 'scaleShuffleTileA': [], 'scaleShuffleTileB': []}, scaleValue_A=1, scaleValue_B=1, initMode_A='DataInitMode(Bounded)', initMode_B='DataInitMode(Bounded)', initMode_C='DataInitMode(Bounded)', workgroupMappingDim=-1, workgroupMappingValue=-1)GEMMSolution(mac_m=128, mac_n=128, mac_k=64, wave_m=32, wave_n=32, wave_k=8, wave_b=1, workgroup_size_x=128, workgroup_size_y=2, workgroupRemapXCC=False, workgroupRemapXCCValue=-1, unroll_x=0, unroll_y=0, load_A='BufferToLDSViaVGPR', load_B='BufferToLDSViaVGPR', storeLDS_D=True, betaInFma=True, scheduler='Priority', schedulerCost='LinearWeighted', prefetch=False, prefetchInFlight=0, prefetchLDSFactor=0, prefetchMixMemOps=False, loadScale_A='BufferToVGPR', loadScale_B='BufferToVGPR', swizzleScale=False, swizzleTileSize={'m': 0, 'k': 0, 'n': 0, 'l': 0}, prefetchScale=False, streamK=True, numWGs=304, streamKTwoTile=False, streamKTwoTileDPFirst=False, architecture={'ArchString': 'gfx942', 'Xnack': False, 'Sramecc': True}, matchMemoryAccess=True, version='')
- SGPR: 80 -> 82 (+2)
	| 3. FloatsGEMM(M: 7680, N: 8448, K: 8448, alpha: 2, beta: 0.5, types: {'type_A': 'half', 'type_B': 'half', 'type_C': 'half', 'type_D': 'half', 'type_acc': 'float', 'trans_A': 'N', 'trans_B': 'T', 'scale_A': 'None', 'scaleType_A': 'None', 'scale_B': 'None', 'scaleType_B': 'None', 'scaleBlockSize': -1, 'scaleSkipPermlane': False, 'scaleShuffleTileA': [], 'scaleShuffleTileB': []}, scaleValue_A: 1, scaleValue_B: 1, initMode_A: DataInitMode(Bounded), initMode_B: DataInitMode(Bounded), initMode_C: DataInitMode(Bounded), workgroupMappingDim: -1, workgroupMappingValue: -1)| GEMM(mac_m: 128, mac_n: 128, mac_k: 64, wave_m: 32, wave_n: 32, wave_k: 8, wave_b: 1, workgroup_size_x: 128, workgroup_size_y: 2, workgroupRemapXCC: False, workgroupRemapXCCValue: -1, unroll_x: 0, unroll_y: 0, load_A: BufferToLDSViaVGPR, load_B: BufferToLDSViaVGPR, storeLDS_D: True, betaInFma: True, scheduler: Priority, schedulerCost: LinearWeighted, prefetch: True, prefetchInFlight: 2, prefetchLDSFactor: 2, prefetchMixMemOps: False, loadScale_A: BufferToVGPR, loadScale_B: BufferToVGPR, swizzleScale: False, swizzleTileSize: {'m': 0, 'k': 0, 'n': 0, 'l': 0}, prefetchScale: False, streamK: True, numWGs: 304, streamKTwoTile: False, streamKTwoTileDPFirst: False, architecture: {'ArchString': 'gfx942', 'Xnack': False, 'Sramecc': True}, matchMemoryAccess: True, version: ) | GEMMProblem(M=7680, N=8448, K=8448, alpha=2, beta=0.5, types={'type_A': 'half', 'type_B': 'half', 'type_C': 'half', 'type_D': 'half', 'type_acc': 'float', 'trans_A': 'N', 'trans_B': 'T', 'scale_A': 'None', 'scaleType_A': 'None', 'scale_B': 'None', 'scaleType_B': 'None', 'scaleBlockSize': -1, 'scaleSkipPermlane': False, 'scaleShuffleTileA': [], 'scaleShuffleTileB': []}, scaleValue_A=1, scaleValue_B=1, initMode_A='DataInitMode(Bounded)', initMode_B='DataInitMode(Bounded)', initMode_C='DataInitMode(Bounded)', workgroupMappingDim=-1, workgroupMappingValue=-1)GEMMSolution(mac_m=128, mac_n=128, mac_k=64, wave_m=32, wave_n=32, wave_k=8, wave_b=1, workgroup_size_x=128, workgroup_size_y=2, workgroupRemapXCC=False, workgroupRemapXCCValue=-1, unroll_x=0, unroll_y=0, load_A='BufferToLDSViaVGPR', load_B='BufferToLDSViaVGPR', storeLDS_D=True, betaInFma=True, scheduler='Priority', schedulerCost='LinearWeighted', prefetch=True, prefetchInFlight=2, prefetchLDSFactor=2, prefetchMixMemOps=False, loadScale_A='BufferToVGPR', loadScale_B='BufferToVGPR', swizzleScale=False, swizzleTileSize={'m': 0, 'k': 0, 'n': 0, 'l': 0}, prefetchScale=False, streamK=True, numWGs=304, streamKTwoTile=False, streamKTwoTileDPFirst=False, architecture={'ArchString': 'gfx942', 'Xnack': False, 'Sramecc': True}, matchMemoryAccess=True, version='')
- SGPR: 80 -> 82 (+2)
	| 3. FloatsGEMM(M: 7680, N: 8448, K: 8448, alpha: 2, beta: 0.5, types: {'type_A': 'half', 'type_B': 'half', 'type_C': 'half', 'type_D': 'half', 'type_acc': 'float', 'trans_A': 'N', 'trans_B': 'T', 'scale_A': 'None', 'scaleType_A': 'None', 'scale_B': 'None', 'scaleType_B': 'None', 'scaleBlockSize': -1, 'scaleSkipPermlane': False, 'scaleShuffleTileA': [], 'scaleShuffleTileB': []}, scaleValue_A: 1, scaleValue_B: 1, initMode_A: DataInitMode(Bounded), initMode_B: DataInitMode(Bounded), initMode_C: DataInitMode(Bounded), workgroupMappingDim: -1, workgroupMappingValue: -1)| GEMM(mac_m: 128, mac_n: 256, mac_k: 16, wave_m: 32, wave_n: 32, wave_k: 8, wave_b: 1, workgroup_size_x: 128, workgroup_size_y: 2, workgroupRemapXCC: False, workgroupRemapXCCValue: -1, unroll_x: 0, unroll_y: 0, load_A: BufferToLDSViaVGPR, load_B: BufferToLDSViaVGPR, storeLDS_D: True, betaInFma: True, scheduler: Priority, schedulerCost: LinearWeighted, prefetch: False, prefetchInFlight: 0, prefetchLDSFactor: 0, prefetchMixMemOps: False, loadScale_A: BufferToVGPR, loadScale_B: BufferToVGPR, swizzleScale: False, swizzleTileSize: {'m': 0, 'k': 0, 'n': 0, 'l': 0}, prefetchScale: False, streamK: True, numWGs: 304, streamKTwoTile: False, streamKTwoTileDPFirst: False, architecture: {'ArchString': 'gfx942', 'Xnack': False, 'Sramecc': True}, matchMemoryAccess: True, version: ) | GEMMProblem(M=7680, N=8448, K=8448, alpha=2, beta=0.5, types={'type_A': 'half', 'type_B': 'half', 'type_C': 'half', 'type_D': 'half', 'type_acc': 'float', 'trans_A': 'N', 'trans_B': 'T', 'scale_A': 'None', 'scaleType_A': 'None', 'scale_B': 'None', 'scaleType_B': 'None', 'scaleBlockSize': -1, 'scaleSkipPermlane': False, 'scaleShuffleTileA': [], 'scaleShuffleTileB': []}, scaleValue_A=1, scaleValue_B=1, initMode_A='DataInitMode(Bounded)', initMode_B='DataInitMode(Bounded)', initMode_C='DataInitMode(Bounded)', workgroupMappingDim=-1, workgroupMappingValue=-1)GEMMSolution(mac_m=128, mac_n=256, mac_k=16, wave_m=32, wave_n=32, wave_k=8, wave_b=1, workgroup_size_x=128, workgroup_size_y=2, workgroupRemapXCC=False, workgroupRemapXCCValue=-1, unroll_x=0, unroll_y=0, load_A='BufferToLDSViaVGPR', load_B='BufferToLDSViaVGPR', storeLDS_D=True, betaInFma=True, scheduler='Priority', schedulerCost='LinearWeighted', prefetch=False, prefetchInFlight=0, prefetchLDSFactor=0, prefetchMixMemOps=False, loadScale_A='BufferToVGPR', loadScale_B='BufferToVGPR', swizzleScale=False, swizzleTileSize={'m': 0, 'k': 0, 'n': 0, 'l': 0}, prefetchScale=False, streamK=True, numWGs=304, streamKTwoTile=False, streamKTwoTileDPFirst=False, architecture={'ArchString': 'gfx942', 'Xnack': False, 'Sramecc': True}, matchMemoryAccess=True, version='')
- SGPR: 84 -> 86 (+2)
	| 3. FloatsGEMM(M: 7680, N: 8448, K: 8448, alpha: 2, beta: 0.5, types: {'type_A': 'half', 'type_B': 'half', 'type_C': 'half', 'type_D': 'half', 'type_acc': 'float', 'trans_A': 'N', 'trans_B': 'T', 'scale_A': 'None', 'scaleType_A': 'None', 'scale_B': 'None', 'scaleType_B': 'None', 'scaleBlockSize': -1, 'scaleSkipPermlane': False, 'scaleShuffleTileA': [], 'scaleShuffleTileB': []}, scaleValue_A: 1, scaleValue_B: 1, initMode_A: DataInitMode(Bounded), initMode_B: DataInitMode(Bounded), initMode_C: DataInitMode(Bounded), workgroupMappingDim: -1, workgroupMappingValue: -1)| GEMM(mac_m: 128, mac_n: 256, mac_k: 16, wave_m: 32, wave_n: 32, wave_k: 8, wave_b: 1, workgroup_size_x: 128, workgroup_size_y: 2, workgroupRemapXCC: False, workgroupRemapXCCValue: -1, unroll_x: 0, unroll_y: 0, load_A: BufferToLDSViaVGPR, load_B: BufferToLDSViaVGPR, storeLDS_D: True, betaInFma: True, scheduler: Priority, schedulerCost: LinearWeighted, prefetch: False, prefetchInFlight: 0, prefetchLDSFactor: 0, prefetchMixMemOps: False, loadScale_A: BufferToVGPR, loadScale_B: BufferToVGPR, swizzleScale: False, swizzleTileSize: {'m': 0, 'k': 0, 'n': 0, 'l': 0}, prefetchScale: False, streamK: True, numWGs: 304, streamKTwoTile: False, streamKTwoTileDPFirst: True, architecture: {'ArchString': 'gfx942', 'Xnack': False, 'Sramecc': True}, matchMemoryAccess: True, version: ) | GEMMProblem(M=7680, N=8448, K=8448, alpha=2, beta=0.5, types={'type_A': 'half', 'type_B': 'half', 'type_C': 'half', 'type_D': 'half', 'type_acc': 'float', 'trans_A': 'N', 'trans_B': 'T', 'scale_A': 'None', 'scaleType_A': 'None', 'scale_B': 'None', 'scaleType_B': 'None', 'scaleBlockSize': -1, 'scaleSkipPermlane': False, 'scaleShuffleTileA': [], 'scaleShuffleTileB': []}, scaleValue_A=1, scaleValue_B=1, initMode_A='DataInitMode(Bounded)', initMode_B='DataInitMode(Bounded)', initMode_C='DataInitMode(Bounded)', workgroupMappingDim=-1, workgroupMappingValue=-1)GEMMSolution(mac_m=128, mac_n=256, mac_k=16, wave_m=32, wave_n=32, wave_k=8, wave_b=1, workgroup_size_x=128, workgroup_size_y=2, workgroupRemapXCC=False, workgroupRemapXCCValue=-1, unroll_x=0, unroll_y=0, load_A='BufferToLDSViaVGPR', load_B='BufferToLDSViaVGPR', storeLDS_D=True, betaInFma=True, scheduler='Priority', schedulerCost='LinearWeighted', prefetch=False, prefetchInFlight=0, prefetchLDSFactor=0, prefetchMixMemOps=False, loadScale_A='BufferToVGPR', loadScale_B='BufferToVGPR', swizzleScale=False, swizzleTileSize={'m': 0, 'k': 0, 'n': 0, 'l': 0}, prefetchScale=False, streamK=True, numWGs=304, streamKTwoTile=False, streamKTwoTileDPFirst=True, architecture={'ArchString': 'gfx942', 'Xnack': False, 'Sramecc': True}, matchMemoryAccess=True, version='')
- SGPR: 84 -> 86 (+2)
	| 3. FloatsGEMM(M: 7680, N: 8448, K: 8448, alpha: 2, beta: 0.5, types: {'type_A': 'half', 'type_B': 'half', 'type_C': 'half', 'type_D': 'half', 'type_acc': 'float', 'trans_A': 'N', 'trans_B': 'T', 'scale_A': 'None', 'scaleType_A': 'None', 'scale_B': 'None', 'scaleType_B': 'None', 'scaleBlockSize': -1, 'scaleSkipPermlane': False, 'scaleShuffleTileA': [], 'scaleShuffleTileB': []}, scaleValue_A: 1, scaleValue_B: 1, initMode_A: DataInitMode(Bounded), initMode_B: DataInitMode(Bounded), initMode_C: DataInitMode(Bounded), workgroupMappingDim: -1, workgroupMappingValue: -1)| GEMM(mac_m: 128, mac_n: 256, mac_k: 16, wave_m: 32, wave_n: 32, wave_k: 8, wave_b: 1, workgroup_size_x: 128, workgroup_size_y: 2, workgroupRemapXCC: False, workgroupRemapXCCValue: -1, unroll_x: 0, unroll_y: 0, load_A: BufferToLDSViaVGPR, load_B: BufferToLDSViaVGPR, storeLDS_D: True, betaInFma: True, scheduler: Priority, schedulerCost: LinearWeighted, prefetch: False, prefetchInFlight: 0, prefetchLDSFactor: 0, prefetchMixMemOps: False, loadScale_A: BufferToVGPR, loadScale_B: BufferToVGPR, swizzleScale: False, swizzleTileSize: {'m': 0, 'k': 0, 'n': 0, 'l': 0}, prefetchScale: False, streamK: True, numWGs: 304, streamKTwoTile: True, streamKTwoTileDPFirst: False, architecture: {'ArchString': 'gfx942', 'Xnack': False, 'Sramecc': True}, matchMemoryAccess: True, version: ) | GEMMProblem(M=7680, N=8448, K=8448, alpha=2, beta=0.5, types={'type_A': 'half', 'type_B': 'half', 'type_C': 'half', 'type_D': 'half', 'type_acc': 'float', 'trans_A': 'N', 'trans_B': 'T', 'scale_A': 'None', 'scaleType_A': 'None', 'scale_B': 'None', 'scaleType_B': 'None', 'scaleBlockSize': -1, 'scaleSkipPermlane': False, 'scaleShuffleTileA': [], 'scaleShuffleTileB': []}, scaleValue_A=1, scaleValue_B=1, initMode_A='DataInitMode(Bounded)', initMode_B='DataInitMode(Bounded)', initMode_C='DataInitMode(Bounded)', workgroupMappingDim=-1, workgroupMappingValue=-1)GEMMSolution(mac_m=128, mac_n=256, mac_k=16, wave_m=32, wave_n=32, wave_k=8, wave_b=1, workgroup_size_x=128, workgroup_size_y=2, workgroupRemapXCC=False, workgroupRemapXCCValue=-1, unroll_x=0, unroll_y=0, load_A='BufferToLDSViaVGPR', load_B='BufferToLDSViaVGPR', storeLDS_D=True, betaInFma=True, scheduler='Priority', schedulerCost='LinearWeighted', prefetch=False, prefetchInFlight=0, prefetchLDSFactor=0, prefetchMixMemOps=False, loadScale_A='BufferToVGPR', loadScale_B='BufferToVGPR', swizzleScale=False, swizzleTileSize={'m': 0, 'k': 0, 'n': 0, 'l': 0}, prefetchScale=False, streamK=True, numWGs=304, streamKTwoTile=True, streamKTwoTileDPFirst=False, architecture={'ArchString': 'gfx942', 'Xnack': False, 'Sramecc': True}, matchMemoryAccess=True, version='')
- SGPR: 80 -> 82 (+2)
	| 3. FloatsGEMM(M: 7680, N: 8448, K: 8448, alpha: 2, beta: 0.5, types: {'type_A': 'half', 'type_B': 'half', 'type_C': 'half', 'type_D': 'half', 'type_acc': 'float', 'trans_A': 'N', 'trans_B': 'T', 'scale_A': 'None', 'scaleType_A': 'None', 'scale_B': 'None', 'scaleType_B': 'None', 'scaleBlockSize': -1, 'scaleSkipPermlane': False, 'scaleShuffleTileA': [], 'scaleShuffleTileB': []}, scaleValue_A: 1, scaleValue_B: 1, initMode_A: DataInitMode(Bounded), initMode_B: DataInitMode(Bounded), initMode_C: DataInitMode(Bounded), workgroupMappingDim: -1, workgroupMappingValue: -1)| GEMM(mac_m: 128, mac_n: 256, mac_k: 16, wave_m: 32, wave_n: 32, wave_k: 8, wave_b: 1, workgroup_size_x: 128, workgroup_size_y: 2, workgroupRemapXCC: False, workgroupRemapXCCValue: -1, unroll_x: 0, unroll_y: 0, load_A: BufferToLDSViaVGPR, load_B: BufferToLDSViaVGPR, storeLDS_D: True, betaInFma: True, scheduler: Priority, schedulerCost: LinearWeighted, prefetch: False, prefetchInFlight: 2, prefetchLDSFactor: 0, prefetchMixMemOps: False, loadScale_A: BufferToVGPR, loadScale_B: BufferToVGPR, swizzleScale: False, swizzleTileSize: {'m': 0, 'k': 0, 'n': 0, 'l': 0}, prefetchScale: False, streamK: True, numWGs: 304, streamKTwoTile: False, streamKTwoTileDPFirst: False, architecture: {'ArchString': 'gfx942', 'Xnack': False, 'Sramecc': True}, matchMemoryAccess: True, version: ) | GEMMProblem(M=7680, N=8448, K=8448, alpha=2, beta=0.5, types={'type_A': 'half', 'type_B': 'half', 'type_C': 'half', 'type_D': 'half', 'type_acc': 'float', 'trans_A': 'N', 'trans_B': 'T', 'scale_A': 'None', 'scaleType_A': 'None', 'scale_B': 'None', 'scaleType_B': 'None', 'scaleBlockSize': -1, 'scaleSkipPermlane': False, 'scaleShuffleTileA': [], 'scaleShuffleTileB': []}, scaleValue_A=1, scaleValue_B=1, initMode_A='DataInitMode(Bounded)', initMode_B='DataInitMode(Bounded)', initMode_C='DataInitMode(Bounded)', workgroupMappingDim=-1, workgroupMappingValue=-1)GEMMSolution(mac_m=128, mac_n=256, mac_k=16, wave_m=32, wave_n=32, wave_k=8, wave_b=1, workgroup_size_x=128, workgroup_size_y=2, workgroupRemapXCC=False, workgroupRemapXCCValue=-1, unroll_x=0, unroll_y=0, load_A='BufferToLDSViaVGPR', load_B='BufferToLDSViaVGPR', storeLDS_D=True, betaInFma=True, scheduler='Priority', schedulerCost='LinearWeighted', prefetch=False, prefetchInFlight=2, prefetchLDSFactor=0, prefetchMixMemOps=False, loadScale_A='BufferToVGPR', loadScale_B='BufferToVGPR', swizzleScale=False, swizzleTileSize={'m': 0, 'k': 0, 'n': 0, 'l': 0}, prefetchScale=False, streamK=True, numWGs=304, streamKTwoTile=False, streamKTwoTileDPFirst=False, architecture={'ArchString': 'gfx942', 'Xnack': False, 'Sramecc': True}, matchMemoryAccess=True, version='')
- SGPR: 84 -> 86 (+2)
	| 3. FloatsGEMM(M: 7680, N: 8448, K: 8448, alpha: 2, beta: 0.5, types: {'type_A': 'half', 'type_B': 'half', 'type_C': 'half', 'type_D': 'half', 'type_acc': 'float', 'trans_A': 'N', 'trans_B': 'T', 'scale_A': 'None', 'scaleType_A': 'None', 'scale_B': 'None', 'scaleType_B': 'None', 'scaleBlockSize': -1, 'scaleSkipPermlane': False, 'scaleShuffleTileA': [], 'scaleShuffleTileB': []}, scaleValue_A: 1, scaleValue_B: 1, initMode_A: DataInitMode(Bounded), initMode_B: DataInitMode(Bounded), initMode_C: DataInitMode(Bounded), workgroupMappingDim: -1, workgroupMappingValue: -1)| GEMM(mac_m: 128, mac_n: 256, mac_k: 16, wave_m: 32, wave_n: 32, wave_k: 8, wave_b: 1, workgroup_size_x: 128, workgroup_size_y: 2, workgroupRemapXCC: False, workgroupRemapXCCValue: -1, unroll_x: 0, unroll_y: 0, load_A: BufferToLDSViaVGPR, load_B: BufferToLDSViaVGPR, storeLDS_D: True, betaInFma: True, scheduler: Priority, schedulerCost: LinearWeighted, prefetch: False, prefetchInFlight: 2, prefetchLDSFactor: 0, prefetchMixMemOps: False, loadScale_A: BufferToVGPR, loadScale_B: BufferToVGPR, swizzleScale: False, swizzleTileSize: {'m': 0, 'k': 0, 'n': 0, 'l': 0}, prefetchScale: False, streamK: True, numWGs: 304, streamKTwoTile: False, streamKTwoTileDPFirst: True, architecture: {'ArchString': 'gfx942', 'Xnack': False, 'Sramecc': True}, matchMemoryAccess: True, version: ) | GEMMProblem(M=7680, N=8448, K=8448, alpha=2, beta=0.5, types={'type_A': 'half', 'type_B': 'half', 'type_C': 'half', 'type_D': 'half', 'type_acc': 'float', 'trans_A': 'N', 'trans_B': 'T', 'scale_A': 'None', 'scaleType_A': 'None', 'scale_B': 'None', 'scaleType_B': 'None', 'scaleBlockSize': -1, 'scaleSkipPermlane': False, 'scaleShuffleTileA': [], 'scaleShuffleTileB': []}, scaleValue_A=1, scaleValue_B=1, initMode_A='DataInitMode(Bounded)', initMode_B='DataInitMode(Bounded)', initMode_C='DataInitMode(Bounded)', workgroupMappingDim=-1, workgroupMappingValue=-1)GEMMSolution(mac_m=128, mac_n=256, mac_k=16, wave_m=32, wave_n=32, wave_k=8, wave_b=1, workgroup_size_x=128, workgroup_size_y=2, workgroupRemapXCC=False, workgroupRemapXCCValue=-1, unroll_x=0, unroll_y=0, load_A='BufferToLDSViaVGPR', load_B='BufferToLDSViaVGPR', storeLDS_D=True, betaInFma=True, scheduler='Priority', schedulerCost='LinearWeighted', prefetch=False, prefetchInFlight=2, prefetchLDSFactor=0, prefetchMixMemOps=False, loadScale_A='BufferToVGPR', loadScale_B='BufferToVGPR', swizzleScale=False, swizzleTileSize={'m': 0, 'k': 0, 'n': 0, 'l': 0}, prefetchScale=False, streamK=True, numWGs=304, streamKTwoTile=False, streamKTwoTileDPFirst=True, architecture={'ArchString': 'gfx942', 'Xnack': False, 'Sramecc': True}, matchMemoryAccess=True, version='')
- SGPR: 84 -> 86 (+2)
	| 3. FloatsGEMM(M: 7680, N: 8448, K: 8448, alpha: 2, beta: 0.5, types: {'type_A': 'half', 'type_B': 'half', 'type_C': 'half', 'type_D': 'half', 'type_acc': 'float', 'trans_A': 'N', 'trans_B': 'T', 'scale_A': 'None', 'scaleType_A': 'None', 'scale_B': 'None', 'scaleType_B': 'None', 'scaleBlockSize': -1, 'scaleSkipPermlane': False, 'scaleShuffleTileA': [], 'scaleShuffleTileB': []}, scaleValue_A: 1, scaleValue_B: 1, initMode_A: DataInitMode(Bounded), initMode_B: DataInitMode(Bounded), initMode_C: DataInitMode(Bounded), workgroupMappingDim: -1, workgroupMappingValue: -1)| GEMM(mac_m: 128, mac_n: 256, mac_k: 16, wave_m: 32, wave_n: 32, wave_k: 8, wave_b: 1, workgroup_size_x: 128, workgroup_size_y: 2, workgroupRemapXCC: False, workgroupRemapXCCValue: -1, unroll_x: 0, unroll_y: 0, load_A: BufferToLDSViaVGPR, load_B: BufferToLDSViaVGPR, storeLDS_D: True, betaInFma: True, scheduler: Priority, schedulerCost: LinearWeighted, prefetch: False, prefetchInFlight: 2, prefetchLDSFactor: 0, prefetchMixMemOps: False, loadScale_A: BufferToVGPR, loadScale_B: BufferToVGPR, swizzleScale: False, swizzleTileSize: {'m': 0, 'k': 0, 'n': 0, 'l': 0}, prefetchScale: False, streamK: True, numWGs: 304, streamKTwoTile: True, streamKTwoTileDPFirst: False, architecture: {'ArchString': 'gfx942', 'Xnack': False, 'Sramecc': True}, matchMemoryAccess: True, version: ) | GEMMProblem(M=7680, N=8448, K=8448, alpha=2, beta=0.5, types={'type_A': 'half', 'type_B': 'half', 'type_C': 'half', 'type_D': 'half', 'type_acc': 'float', 'trans_A': 'N', 'trans_B': 'T', 'scale_A': 'None', 'scaleType_A': 'None', 'scale_B': 'None', 'scaleType_B': 'None', 'scaleBlockSize': -1, 'scaleSkipPermlane': False, 'scaleShuffleTileA': [], 'scaleShuffleTileB': []}, scaleValue_A=1, scaleValue_B=1, initMode_A='DataInitMode(Bounded)', initMode_B='DataInitMode(Bounded)', initMode_C='DataInitMode(Bounded)', workgroupMappingDim=-1, workgroupMappingValue=-1)GEMMSolution(mac_m=128, mac_n=256, mac_k=16, wave_m=32, wave_n=32, wave_k=8, wave_b=1, workgroup_size_x=128, workgroup_size_y=2, workgroupRemapXCC=False, workgroupRemapXCCValue=-1, unroll_x=0, unroll_y=0, load_A='BufferToLDSViaVGPR', load_B='BufferToLDSViaVGPR', storeLDS_D=True, betaInFma=True, scheduler='Priority', schedulerCost='LinearWeighted', prefetch=False, prefetchInFlight=2, prefetchLDSFactor=0, prefetchMixMemOps=False, loadScale_A='BufferToVGPR', loadScale_B='BufferToVGPR', swizzleScale=False, swizzleTileSize={'m': 0, 'k': 0, 'n': 0, 'l': 0}, prefetchScale=False, streamK=True, numWGs=304, streamKTwoTile=True, streamKTwoTileDPFirst=False, architecture={'ArchString': 'gfx942', 'Xnack': False, 'Sramecc': True}, matchMemoryAccess=True, version='')
- SGPR: 80 -> 82 (+2)

Results truncated, see full report in workspace

Links

memmett · 2025-12-09T23:07:37Z

There was a test that Nathan enabled which uncovered a problem with multiple fixups -- I see you have a fix for that in this PR. Can you enable the test that Nathan found?

…braries into users/liu-yiqian/reset-flag

liu-yiqian · 2025-12-10T17:34:26Z

There was a test that Nathan enabled which uncovered a problem with multiple fixups -- I see you have a fix for that in this PR. Can you enable the test that Nathan found?

Commit a2979c0 adds this test. I also have a following up PR to add StreamK suite.

math-ci · 2025-12-10T18:33:01Z

Performance Report for gfx950-rocroller

Results

Details

@@            Significant (p-val <0.05) Performance Diffs            @@
====================================================================================================
+   0.10% | p=6.2764e-06 
	| 3. FloatsGEMM(M: 4096, N: 4096, K: 32768, alpha: 2, beta: 0, types: {'type_A': 'fp4', 'type_B': 'fp4', 'type_C': 'half', 'type_D': 'half', 'type_acc': 'float', 'trans_A': 'T', 'trans_B': 'N', 'scale_A': 'Separate', 'scaleType_A': 'E8M0', 'scale_B': 'Separate', 'scaleType_B': 'E8M0', 'scaleBlockSize': 32, 'scaleSkipPermlane': False, 'scaleShuffleTileA': [], 'scaleShuffleTileB': []}, scaleValue_A: 1, scaleValue_B: 1, initMode_A: DataInitMode(Bounded), initMode_B: DataInitMode(Bounded), initMode_C: DataInitMode(Bounded), workgroupMappingDim: -1, workgroupMappingValue: -1)| GEMM(mac_m: 128, mac_n: 128, mac_k: 256, wave_m: 32, wave_n: 32, wave_k: 64, wave_b: 1, workgroup_size_x: 128, workgroup_size_y: 2, workgroupRemapXCC: False, workgroupRemapXCCValue: -1, unroll_x: 0, unroll_y: 0, load_A: BufferToLDS, load_B: BufferToLDS, storeLDS_D: False, betaInFma: True, scheduler: Priority, schedulerCost: LinearWeighted, prefetch: True, prefetchInFlight: 4, prefetchLDSFactor: 1, prefetchMixMemOps: True, loadScale_A: BufferToVGPR, loadScale_B: BufferToVGPR, swizzleScale: True, swizzleTileSize: {'m': 32, 'k': 8, 'n': 32, 'l': 8}, prefetchScale: True, streamK: False, numWGs: 0, streamKTwoTile: False, streamKTwoTileDPFirst: False, architecture: {'ArchString': 'gfx950', 'Xnack': False, 'Sramecc': True}, matchMemoryAccess: True, version: ) | GEMMProblem(M=4096, N=4096, K=32768, alpha=2, beta=0, types={'type_A': 'fp4', 'type_B': 'fp4', 'type_C': 'half', 'type_D': 'half', 'type_acc': 'float', 'trans_A': 'T', 'trans_B': 'N', 'scale_A': 'Separate', 'scaleType_A': 'E8M0', 'scale_B': 'Separate', 'scaleType_B': 'E8M0', 'scaleBlockSize': 32, 'scaleSkipPermlane': False, 'scaleShuffleTileA': [], 'scaleShuffleTileB': []}, scaleValue_A=1, scaleValue_B=1, initMode_A='DataInitMode(Bounded)', initMode_B='DataInitMode(Bounded)', initMode_C='DataInitMode(Bounded)', workgroupMappingDim=-1, workgroupMappingValue=-1)GEMMSolution(mac_m=128, mac_n=128, mac_k=256, wave_m=32, wave_n=32, wave_k=64, wave_b=1, workgroup_size_x=128, workgroup_size_y=2, workgroupRemapXCC=False, workgroupRemapXCCValue=-1, unroll_x=0, unroll_y=0, load_A='BufferToLDS', load_B='BufferToLDS', storeLDS_D=False, betaInFma=True, scheduler='Priority', schedulerCost='LinearWeighted', prefetch=True, prefetchInFlight=4, prefetchLDSFactor=1, prefetchMixMemOps=True, loadScale_A='BufferToVGPR', loadScale_B='BufferToVGPR', swizzleScale=True, swizzleTileSize={'m': 32, 'k': 8, 'n': 32, 'l': 8}, prefetchScale=True, streamK=False, numWGs=0, streamKTwoTile=False, streamKTwoTileDPFirst=False, architecture={'ArchString': 'gfx950', 'Xnack': False, 'Sramecc': True}, matchMemoryAccess=True, version='')
-   0.40% | p=1.0272e-07 
	| 3. FloatsGEMM(M: 4096, N: 4096, K: 32768, alpha: 2, beta: 0, types: {'type_A': 'fp4', 'type_B': 'fp4', 'type_C': 'half', 'type_D': 'half', 'type_acc': 'float', 'trans_A': 'T', 'trans_B': 'N', 'scale_A': 'Separate', 'scaleType_A': 'E8M0', 'scale_B': 'Separate', 'scaleType_B': 'E8M0', 'scaleBlockSize': 32, 'scaleSkipPermlane': False, 'scaleShuffleTileA': [], 'scaleShuffleTileB': []}, scaleValue_A: 1, scaleValue_B: 1, initMode_A: DataInitMode(Bounded), initMode_B: DataInitMode(Bounded), initMode_C: DataInitMode(Bounded), workgroupMappingDim: -1, workgroupMappingValue: -1)| GEMM(mac_m: 256, mac_n: 256, mac_k: 128, wave_m: 16, wave_n: 16, wave_k: 128, wave_b: 1, workgroup_size_x: 128, workgroup_size_y: 2, workgroupRemapXCC: False, workgroupRemapXCCValue: -1, unroll_x: 0, unroll_y: 0, load_A: BufferToLDS, load_B: BufferToLDS, storeLDS_D: False, betaInFma: True, scheduler: Priority, schedulerCost: LinearWeightedSimple, prefetch: True, prefetchInFlight: 4, prefetchLDSFactor: 1, prefetchMixMemOps: True, loadScale_A: BufferToVGPR, loadScale_B: BufferToVGPR, swizzleScale: True, swizzleTileSize: {'m': 64, 'k': 16, 'n': 64, 'l': 16}, prefetchScale: True, streamK: False, numWGs: 0, streamKTwoTile: False, streamKTwoTileDPFirst: False, architecture: {'ArchString': 'gfx950', 'Xnack': False, 'Sramecc': True}, matchMemoryAccess: True, version: ) | GEMMProblem(M=4096, N=4096, K=32768, alpha=2, beta=0, types={'type_A': 'fp4', 'type_B': 'fp4', 'type_C': 'half', 'type_D': 'half', 'type_acc': 'float', 'trans_A': 'T', 'trans_B': 'N', 'scale_A': 'Separate', 'scaleType_A': 'E8M0', 'scale_B': 'Separate', 'scaleType_B': 'E8M0', 'scaleBlockSize': 32, 'scaleSkipPermlane': False, 'scaleShuffleTileA': [], 'scaleShuffleTileB': []}, scaleValue_A=1, scaleValue_B=1, initMode_A='DataInitMode(Bounded)', initMode_B='DataInitMode(Bounded)', initMode_C='DataInitMode(Bounded)', workgroupMappingDim=-1, workgroupMappingValue=-1)GEMMSolution(mac_m=256, mac_n=256, mac_k=128, wave_m=16, wave_n=16, wave_k=128, wave_b=1, workgroup_size_x=128, workgroup_size_y=2, workgroupRemapXCC=False, workgroupRemapXCCValue=-1, unroll_x=0, unroll_y=0, load_A='BufferToLDS', load_B='BufferToLDS', storeLDS_D=False, betaInFma=True, scheduler='Priority', schedulerCost='LinearWeightedSimple', prefetch=True, prefetchInFlight=4, prefetchLDSFactor=1, prefetchMixMemOps=True, loadScale_A='BufferToVGPR', loadScale_B='BufferToVGPR', swizzleScale=True, swizzleTileSize={'m': 64, 'k': 16, 'n': 64, 'l': 16}, prefetchScale=True, streamK=False, numWGs=0, streamKTwoTile=False, streamKTwoTileDPFirst=False, architecture={'ArchString': 'gfx950', 'Xnack': False, 'Sramecc': True}, matchMemoryAccess=True, version='')
-   1.06% | p=8.4176e-102 
	| 3. FloatsGEMM(M: 4096, N: 4096, K: 32768, alpha: 2, beta: 0, types: {'type_A': 'fp4', 'type_B': 'fp4', 'type_C': 'half', 'type_D': 'half', 'type_acc': 'float', 'trans_A': 'T', 'trans_B': 'N', 'scale_A': 'Separate', 'scaleType_A': 'E8M0', 'scale_B': 'Separate', 'scaleType_B': 'E8M0', 'scaleBlockSize': 32, 'scaleSkipPermlane': False, 'scaleShuffleTileA': [], 'scaleShuffleTileB': []}, scaleValue_A: 1, scaleValue_B: 1, initMode_A: DataInitMode(Bounded), initMode_B: DataInitMode(Bounded), initMode_C: DataInitMode(Bounded), workgroupMappingDim: -1, workgroupMappingValue: -1)| GEMM(mac_m: 256, mac_n: 256, mac_k: 128, wave_m: 32, wave_n: 32, wave_k: 64, wave_b: 1, workgroup_size_x: 128, workgroup_size_y: 2, workgroupRemapXCC: False, workgroupRemapXCCValue: -1, unroll_x: 0, unroll_y: 0, load_A: BufferToLDS, load_B: BufferToLDS, storeLDS_D: False, betaInFma: True, scheduler: Priority, schedulerCost: LinearWeighted, prefetch: True, prefetchInFlight: 4, prefetchLDSFactor: 1, prefetchMixMemOps: True, loadScale_A: BufferToVGPR, loadScale_B: BufferToVGPR, swizzleScale: True, swizzleTileSize: {'m': 64, 'k': 16, 'n': 64, 'l': 16}, prefetchScale: True, streamK: False, numWGs: 0, streamKTwoTile: False, streamKTwoTileDPFirst: False, architecture: {'ArchString': 'gfx950', 'Xnack': False, 'Sramecc': True}, matchMemoryAccess: True, version: ) | GEMMProblem(M=4096, N=4096, K=32768, alpha=2, beta=0, types={'type_A': 'fp4', 'type_B': 'fp4', 'type_C': 'half', 'type_D': 'half', 'type_acc': 'float', 'trans_A': 'T', 'trans_B': 'N', 'scale_A': 'Separate', 'scaleType_A': 'E8M0', 'scale_B': 'Separate', 'scaleType_B': 'E8M0', 'scaleBlockSize': 32, 'scaleSkipPermlane': False, 'scaleShuffleTileA': [], 'scaleShuffleTileB': []}, scaleValue_A=1, scaleValue_B=1, initMode_A='DataInitMode(Bounded)', initMode_B='DataInitMode(Bounded)', initMode_C='DataInitMode(Bounded)', workgroupMappingDim=-1, workgroupMappingValue=-1)GEMMSolution(mac_m=256, mac_n=256, mac_k=128, wave_m=32, wave_n=32, wave_k=64, wave_b=1, workgroup_size_x=128, workgroup_size_y=2, workgroupRemapXCC=False, workgroupRemapXCCValue=-1, unroll_x=0, unroll_y=0, load_A='BufferToLDS', load_B='BufferToLDS', storeLDS_D=False, betaInFma=True, scheduler='Priority', schedulerCost='LinearWeighted', prefetch=True, prefetchInFlight=4, prefetchLDSFactor=1, prefetchMixMemOps=True, loadScale_A='BufferToVGPR', loadScale_B='BufferToVGPR', swizzleScale=True, swizzleTileSize={'m': 64, 'k': 16, 'n': 64, 'l': 16}, prefetchScale=True, streamK=False, numWGs=0, streamKTwoTile=False, streamKTwoTileDPFirst=False, architecture={'ArchString': 'gfx950', 'Xnack': False, 'Sramecc': True}, matchMemoryAccess=True, version='')
-   0.36% | p=1.2620e-51 
	| 3. FloatsGEMM(M: 4096, N: 4096, K: 32768, alpha: 2, beta: 0, types: {'type_A': 'fp4', 'type_B': 'fp4', 'type_C': 'half', 'type_D': 'half', 'type_acc': 'float', 'trans_A': 'T', 'trans_B': 'N', 'scale_A': 'Separate', 'scaleType_A': 'E8M0', 'scale_B': 'Separate', 'scaleType_B': 'E8M0', 'scaleBlockSize': 32, 'scaleSkipPermlane': False, 'scaleShuffleTileA': [], 'scaleShuffleTileB': []}, scaleValue_A: 1, scaleValue_B: 1, initMode_A: DataInitMode(Bounded), initMode_B: DataInitMode(Bounded), initMode_C: DataInitMode(Bounded), workgroupMappingDim: 0, workgroupMappingValue: -1)| GEMM(mac_m: 128, mac_n: 128, mac_k: 256, wave_m: 32, wave_n: 32, wave_k: 64, wave_b: 1, workgroup_size_x: 128, workgroup_size_y: 2, workgroupRemapXCC: False, workgroupRemapXCCValue: -1, unroll_x: 0, unroll_y: 0, load_A: BufferToLDS, load_B: BufferToLDS, storeLDS_D: False, betaInFma: True, scheduler: Priority, schedulerCost: LinearWeighted, prefetch: True, prefetchInFlight: 4, prefetchLDSFactor: 1, prefetchMixMemOps: True, loadScale_A: BufferToVGPR, loadScale_B: BufferToVGPR, swizzleScale: True, swizzleTileSize: {'m': 32, 'k': 8, 'n': 32, 'l': 8}, prefetchScale: True, streamK: False, numWGs: 0, streamKTwoTile: False, streamKTwoTileDPFirst: False, architecture: {'ArchString': 'gfx950', 'Xnack': False, 'Sramecc': True}, matchMemoryAccess: True, version: ) | GEMMProblem(M=4096, N=4096, K=32768, alpha=2, beta=0, types={'type_A': 'fp4', 'type_B': 'fp4', 'type_C': 'half', 'type_D': 'half', 'type_acc': 'float', 'trans_A': 'T', 'trans_B': 'N', 'scale_A': 'Separate', 'scaleType_A': 'E8M0', 'scale_B': 'Separate', 'scaleType_B': 'E8M0', 'scaleBlockSize': 32, 'scaleSkipPermlane': False, 'scaleShuffleTileA': [], 'scaleShuffleTileB': []}, scaleValue_A=1, scaleValue_B=1, initMode_A='DataInitMode(Bounded)', initMode_B='DataInitMode(Bounded)', initMode_C='DataInitMode(Bounded)', workgroupMappingDim=0, workgroupMappingValue=-1)GEMMSolution(mac_m=128, mac_n=128, mac_k=256, wave_m=32, wave_n=32, wave_k=64, wave_b=1, workgroup_size_x=128, workgroup_size_y=2, workgroupRemapXCC=False, workgroupRemapXCCValue=-1, unroll_x=0, unroll_y=0, load_A='BufferToLDS', load_B='BufferToLDS', storeLDS_D=False, betaInFma=True, scheduler='Priority', schedulerCost='LinearWeighted', prefetch=True, prefetchInFlight=4, prefetchLDSFactor=1, prefetchMixMemOps=True, loadScale_A='BufferToVGPR', loadScale_B='BufferToVGPR', swizzleScale=True, swizzleTileSize={'m': 32, 'k': 8, 'n': 32, 'l': 8}, prefetchScale=True, streamK=False, numWGs=0, streamKTwoTile=False, streamKTwoTileDPFirst=False, architecture={'ArchString': 'gfx950', 'Xnack': False, 'Sramecc': True}, matchMemoryAccess=True, version='')
-   0.04% | p=3.9597e-02 
	| 3. FloatsGEMM(M: 4096, N: 4096, K: 32768, alpha: 2, beta: 0, types: {'type_A': 'fp4', 'type_B': 'fp4', 'type_C': 'half', 'type_D': 'half', 'type_acc': 'float', 'trans_A': 'T', 'trans_B': 'N', 'scale_A': 'Separate', 'scaleType_A': 'E8M0', 'scale_B': 'Separate', 'scaleType_B': 'E8M0', 'scaleBlockSize': 32, 'scaleSkipPermlane': False, 'scaleShuffleTileA': [], 'scaleShuffleTileB': []}, scaleValue_A: 1, scaleValue_B: 1, initMode_A: DataInitMode(Bounded), initMode_B: DataInitMode(Bounded), initMode_C: DataInitMode(Bounded), workgroupMappingDim: 0, workgroupMappingValue: -1)| GEMM(mac_m: 256, mac_n: 256, mac_k: 128, wave_m: 32, wave_n: 32, wave_k: 64, wave_b: 1, workgroup_size_x: 128, workgroup_size_y: 2, workgroupRemapXCC: False, workgroupRemapXCCValue: -1, unroll_x: 0, unroll_y: 0, load_A: BufferToLDS, load_B: BufferToLDS, storeLDS_D: False, betaInFma: True, scheduler: Priority, schedulerCost: LinearWeighted, prefetch: True, prefetchInFlight: 4, prefetchLDSFactor: 1, prefetchMixMemOps: True, loadScale_A: BufferToVGPR, loadScale_B: BufferToVGPR, swizzleScale: True, swizzleTileSize: {'m': 64, 'k': 16, 'n': 64, 'l': 16}, prefetchScale: True, streamK: False, numWGs: 0, streamKTwoTile: False, streamKTwoTileDPFirst: False, architecture: {'ArchString': 'gfx950', 'Xnack': False, 'Sramecc': True}, matchMemoryAccess: True, version: ) | GEMMProblem(M=4096, N=4096, K=32768, alpha=2, beta=0, types={'type_A': 'fp4', 'type_B': 'fp4', 'type_C': 'half', 'type_D': 'half', 'type_acc': 'float', 'trans_A': 'T', 'trans_B': 'N', 'scale_A': 'Separate', 'scaleType_A': 'E8M0', 'scale_B': 'Separate', 'scaleType_B': 'E8M0', 'scaleBlockSize': 32, 'scaleSkipPermlane': False, 'scaleShuffleTileA': [], 'scaleShuffleTileB': []}, scaleValue_A=1, scaleValue_B=1, initMode_A='DataInitMode(Bounded)', initMode_B='DataInitMode(Bounded)', initMode_C='DataInitMode(Bounded)', workgroupMappingDim=0, workgroupMappingValue=-1)GEMMSolution(mac_m=256, mac_n=256, mac_k=128, wave_m=32, wave_n=32, wave_k=64, wave_b=1, workgroup_size_x=128, workgroup_size_y=2, workgroupRemapXCC=False, workgroupRemapXCCValue=-1, unroll_x=0, unroll_y=0, load_A='BufferToLDS', load_B='BufferToLDS', storeLDS_D=False, betaInFma=True, scheduler='Priority', schedulerCost='LinearWeighted', prefetch=True, prefetchInFlight=4, prefetchLDSFactor=1, prefetchMixMemOps=True, loadScale_A='BufferToVGPR', loadScale_B='BufferToVGPR', swizzleScale=True, swizzleTileSize={'m': 64, 'k': 16, 'n': 64, 'l': 16}, prefetchScale=True, streamK=False, numWGs=0, streamKTwoTile=False, streamKTwoTileDPFirst=False, architecture={'ArchString': 'gfx950', 'Xnack': False, 'Sramecc': True}, matchMemoryAccess=True, version='')
-   0.56% | p=4.8146e-92 
	| 3. FloatsGEMM(M: 4096, N: 4096, K: 32768, alpha: 2, beta: 0, types: {'type_A': 'fp4', 'type_B': 'fp4', 'type_C': 'half', 'type_D': 'half', 'type_acc': 'float', 'trans_A': 'T', 'trans_B': 'N', 'scale_A': 'Separate', 'scaleType_A': 'E8M0', 'scale_B': 'Separate', 'scaleType_B': 'E8M0', 'scaleBlockSize': 32, 'scaleSkipPermlane': True, 'scaleShuffleTileA': [32, 8, 2], 'scaleShuffleTileB': [32, 8, 2]}, scaleValue_A: 1, scaleValue_B: 1, initMode_A: DataInitMode(Bounded), initMode_B: DataInitMode(Bounded), initMode_C: DataInitMode(Bounded), workgroupMappingDim: 0, workgroupMappingValue: -1)| GEMM(mac_m: 128, mac_n: 128, mac_k: 256, wave_m: 32, wave_n: 32, wave_k: 64, wave_b: 1, workgroup_size_x: 128, workgroup_size_y: 2, workgroupRemapXCC: False, workgroupRemapXCCValue: -1, unroll_x: 0, unroll_y: 0, load_A: BufferToLDS, load_B: BufferToLDS, storeLDS_D: False, betaInFma: True, scheduler: Priority, schedulerCost: LinearWeighted, prefetch: True, prefetchInFlight: 4, prefetchLDSFactor: 1, prefetchMixMemOps: True, loadScale_A: BufferToVGPR, loadScale_B: BufferToVGPR, swizzleScale: True, swizzleTileSize: {'m': 32, 'k': 8, 'n': 32, 'l': 8}, prefetchScale: True, streamK: False, numWGs: 0, streamKTwoTile: False, streamKTwoTileDPFirst: False, architecture: {'ArchString': 'gfx950', 'Xnack': False, 'Sramecc': True}, matchMemoryAccess: True, version: ) | GEMMProblem(M=4096, N=4096, K=32768, alpha=2, beta=0, types={'type_A': 'fp4', 'type_B': 'fp4', 'type_C': 'half', 'type_D': 'half', 'type_acc': 'float', 'trans_A': 'T', 'trans_B': 'N', 'scale_A': 'Separate', 'scaleType_A': 'E8M0', 'scale_B': 'Separate', 'scaleType_B': 'E8M0', 'scaleBlockSize': 32, 'scaleSkipPermlane': True, 'scaleShuffleTileA': [32, 8, 2], 'scaleShuffleTileB': [32, 8, 2]}, scaleValue_A=1, scaleValue_B=1, initMode_A='DataInitMode(Bounded)', initMode_B='DataInitMode(Bounded)', initMode_C='DataInitMode(Bounded)', workgroupMappingDim=0, workgroupMappingValue=-1)GEMMSolution(mac_m=128, mac_n=128, mac_k=256, wave_m=32, wave_n=32, wave_k=64, wave_b=1, workgroup_size_x=128, workgroup_size_y=2, workgroupRemapXCC=False, workgroupRemapXCCValue=-1, unroll_x=0, unroll_y=0, load_A='BufferToLDS', load_B='BufferToLDS', storeLDS_D=False, betaInFma=True, scheduler='Priority', schedulerCost='LinearWeighted', prefetch=True, prefetchInFlight=4, prefetchLDSFactor=1, prefetchMixMemOps=True, loadScale_A='BufferToVGPR', loadScale_B='BufferToVGPR', swizzleScale=True, swizzleTileSize={'m': 32, 'k': 8, 'n': 32, 'l': 8}, prefetchScale=True, streamK=False, numWGs=0, streamKTwoTile=False, streamKTwoTileDPFirst=False, architecture={'ArchString': 'gfx950', 'Xnack': False, 'Sramecc': True}, matchMemoryAccess=True, version='')
+   0.41% | p=1.3500e-08 
	| 3. FloatsGEMM(M: 4096, N: 4096, K: 32768, alpha: 2, beta: 0, types: {'type_A': 'fp4', 'type_B': 'fp4', 'type_C': 'half', 'type_D': 'half', 'type_acc': 'float', 'trans_A': 'T', 'trans_B': 'N', 'scale_A': 'Separate', 'scaleType_A': 'E8M0', 'scale_B': 'Separate', 'scaleType_B': 'E8M0', 'scaleBlockSize': 32, 'scaleSkipPermlane': True, 'scaleShuffleTileA': [32, 8, 4], 'scaleShuffleTileB': [32, 8, 4]}, scaleValue_A: 1, scaleValue_B: 1, initMode_A: DataInitMode(Bounded), initMode_B: DataInitMode(Bounded), initMode_C: DataInitMode(Bounded), workgroupMappingDim: 0, workgroupMappingValue: -1)| GEMM(mac_m: 128, mac_n: 128, mac_k: 256, wave_m: 16, wave_n: 16, wave_k: 128, wave_b: 1, workgroup_size_x: 128, workgroup_size_y: 2, workgroupRemapXCC: False, workgroupRemapXCCValue: -1, unroll_x: 0, unroll_y: 0, load_A: BufferToLDS, load_B: BufferToLDS, storeLDS_D: False, betaInFma: True, scheduler: Priority, schedulerCost: LinearWeightedSimple, prefetch: True, prefetchInFlight: 2, prefetchLDSFactor: 1, prefetchMixMemOps: True, loadScale_A: BufferToVGPR, loadScale_B: BufferToVGPR, swizzleScale: True, swizzleTileSize: {'m': 32, 'k': 8, 'n': 32, 'l': 8}, prefetchScale: True, streamK: False, numWGs: 0, streamKTwoTile: False, streamKTwoTileDPFirst: False, architecture: {'ArchString': 'gfx950', 'Xnack': False, 'Sramecc': True}, matchMemoryAccess: True, version: ) | GEMMProblem(M=4096, N=4096, K=32768, alpha=2, beta=0, types={'type_A': 'fp4', 'type_B': 'fp4', 'type_C': 'half', 'type_D': 'half', 'type_acc': 'float', 'trans_A': 'T', 'trans_B': 'N', 'scale_A': 'Separate', 'scaleType_A': 'E8M0', 'scale_B': 'Separate', 'scaleType_B': 'E8M0', 'scaleBlockSize': 32, 'scaleSkipPermlane': True, 'scaleShuffleTileA': [32, 8, 4], 'scaleShuffleTileB': [32, 8, 4]}, scaleValue_A=1, scaleValue_B=1, initMode_A='DataInitMode(Bounded)', initMode_B='DataInitMode(Bounded)', initMode_C='DataInitMode(Bounded)', workgroupMappingDim=0, workgroupMappingValue=-1)GEMMSolution(mac_m=128, mac_n=128, mac_k=256, wave_m=16, wave_n=16, wave_k=128, wave_b=1, workgroup_size_x=128, workgroup_size_y=2, workgroupRemapXCC=False, workgroupRemapXCCValue=-1, unroll_x=0, unroll_y=0, load_A='BufferToLDS', load_B='BufferToLDS', storeLDS_D=False, betaInFma=True, scheduler='Priority', schedulerCost='LinearWeightedSimple', prefetch=True, prefetchInFlight=2, prefetchLDSFactor=1, prefetchMixMemOps=True, loadScale_A='BufferToVGPR', loadScale_B='BufferToVGPR', swizzleScale=True, swizzleTileSize={'m': 32, 'k': 8, 'n': 32, 'l': 8}, prefetchScale=True, streamK=False, numWGs=0, streamKTwoTile=False, streamKTwoTileDPFirst=False, architecture={'ArchString': 'gfx950', 'Xnack': False, 'Sramecc': True}, matchMemoryAccess=True, version='')
-   0.22% | p=8.1060e-24 
	| 3. FloatsGEMM(M: 4096, N: 4096, K: 32768, alpha: 2, beta: 0, types: {'type_A': 'fp4', 'type_B': 'fp4', 'type_C': 'half', 'type_D': 'half', 'type_acc': 'float', 'trans_A': 'T', 'trans_B': 'N', 'scale_A': 'Separate', 'scaleType_A': 'E8M0', 'scale_B': 'Separate', 'scaleType_B': 'E8M0', 'scaleBlockSize': 32, 'scaleSkipPermlane': True, 'scaleShuffleTileA': [64, 4, 2], 'scaleShuffleTileB': [64, 4, 2]}, scaleValue_A: 1, scaleValue_B: 1, initMode_A: DataInitMode(Bounded), initMode_B: DataInitMode(Bounded), initMode_C: DataInitMode(Bounded), workgroupMappingDim: 0, workgroupMappingValue: -1)| GEMM(mac_m: 256, mac_n: 256, mac_k: 128, wave_m: 32, wave_n: 32, wave_k: 64, wave_b: 1, workgroup_size_x: 128, workgroup_size_y: 2, workgroupRemapXCC: False, workgroupRemapXCCValue: -1, unroll_x: 0, unroll_y: 0, load_A: BufferToLDS, load_B: BufferToLDS, storeLDS_D: False, betaInFma: True, scheduler: Priority, schedulerCost: LinearWeighted, prefetch: True, prefetchInFlight: 4, prefetchLDSFactor: 1, prefetchMixMemOps: True, loadScale_A: BufferToVGPR, loadScale_B: BufferToVGPR, swizzleScale: True, swizzleTileSize: {'m': 64, 'k': 16, 'n': 64, 'l': 16}, prefetchScale: True, streamK: False, numWGs: 0, streamKTwoTile: False, streamKTwoTileDPFirst: False, architecture: {'ArchString': 'gfx950', 'Xnack': False, 'Sramecc': True}, matchMemoryAccess: True, version: ) | GEMMProblem(M=4096, N=4096, K=32768, alpha=2, beta=0, types={'type_A': 'fp4', 'type_B': 'fp4', 'type_C': 'half', 'type_D': 'half', 'type_acc': 'float', 'trans_A': 'T', 'trans_B': 'N', 'scale_A': 'Separate', 'scaleType_A': 'E8M0', 'scale_B': 'Separate', 'scaleType_B': 'E8M0', 'scaleBlockSize': 32, 'scaleSkipPermlane': True, 'scaleShuffleTileA': [64, 4, 2], 'scaleShuffleTileB': [64, 4, 2]}, scaleValue_A=1, scaleValue_B=1, initMode_A='DataInitMode(Bounded)', initMode_B='DataInitMode(Bounded)', initMode_C='DataInitMode(Bounded)', workgroupMappingDim=0, workgroupMappingValue=-1)GEMMSolution(mac_m=256, mac_n=256, mac_k=128, wave_m=32, wave_n=32, wave_k=64, wave_b=1, workgroup_size_x=128, workgroup_size_y=2, workgroupRemapXCC=False, workgroupRemapXCCValue=-1, unroll_x=0, unroll_y=0, load_A='BufferToLDS', load_B='BufferToLDS', storeLDS_D=False, betaInFma=True, scheduler='Priority', schedulerCost='LinearWeighted', prefetch=True, prefetchInFlight=4, prefetchLDSFactor=1, prefetchMixMemOps=True, loadScale_A='BufferToVGPR', loadScale_B='BufferToVGPR', swizzleScale=True, swizzleTileSize={'m': 64, 'k': 16, 'n': 64, 'l': 16}, prefetchScale=True, streamK=False, numWGs=0, streamKTwoTile=False, streamKTwoTileDPFirst=False, architecture={'ArchString': 'gfx950', 'Xnack': False, 'Sramecc': True}, matchMemoryAccess=True, version='')
+   0.15% | p=7.7441e-06 
	| 3. FloatsGEMM(M: 4096, N: 4096, K: 32768, alpha: 2, beta: 0, types: {'type_A': 'fp8', 'type_B': 'fp8', 'type_C': 'half', 'type_D': 'half', 'type_acc': 'float', 'trans_A': 'T', 'trans_B': 'N', 'scale_A': 'None', 'scaleType_A': 'None', 'scale_B': 'None', 'scaleType_B': 'None', 'scaleBlockSize': -1, 'scaleSkipPermlane': False, 'scaleShuffleTileA': [], 'scaleShuffleTileB': []}, scaleValue_A: 1, scaleValue_B: 1, initMode_A: DataInitMode(Bounded), initMode_B: DataInitMode(Bounded), initMode_C: DataInitMode(Bounded), workgroupMappingDim: -1, workgroupMappingValue: -1)| GEMM(mac_m: 128, mac_n: 128, mac_k: 128, wave_m: 16, wave_n: 16, wave_k: 128, wave_b: 1, workgroup_size_x: 128, workgroup_size_y: 2, workgroupRemapXCC: False, workgroupRemapXCCValue: -1, unroll_x: 0, unroll_y: 0, load_A: BufferToLDS, load_B: BufferToLDS, storeLDS_D: False, betaInFma: True, scheduler: Priority, schedulerCost: LinearWeighted, prefetch: True, prefetchInFlight: 4, prefetchLDSFactor: 1, prefetchMixMemOps: True, loadScale_A: BufferToVGPR, loadScale_B: BufferToVGPR, swizzleScale: False, swizzleTileSize: {'m': 0, 'k': 0, 'n': 0, 'l': 0}, prefetchScale: False, streamK: False, numWGs: 0, streamKTwoTile: False, streamKTwoTileDPFirst: False, architecture: {'ArchString': 'gfx950', 'Xnack': False, 'Sramecc': True}, matchMemoryAccess: True, version: ) | GEMMProblem(M=4096, N=4096, K=32768, alpha=2, beta=0, types={'type_A': 'fp8', 'type_B': 'fp8', 'type_C': 'half', 'type_D': 'half', 'type_acc': 'float', 'trans_A': 'T', 'trans_B': 'N', 'scale_A': 'None', 'scaleType_A': 'None', 'scale_B': 'None', 'scaleType_B': 'None', 'scaleBlockSize': -1, 'scaleSkipPermlane': False, 'scaleShuffleTileA': [], 'scaleShuffleTileB': []}, scaleValue_A=1, scaleValue_B=1, initMode_A='DataInitMode(Bounded)', initMode_B='DataInitMode(Bounded)', initMode_C='DataInitMode(Bounded)', workgroupMappingDim=-1, workgroupMappingValue=-1)GEMMSolution(mac_m=128, mac_n=128, mac_k=128, wave_m=16, wave_n=16, wave_k=128, wave_b=1, workgroup_size_x=128, workgroup_size_y=2, workgroupRemapXCC=False, workgroupRemapXCCValue=-1, unroll_x=0, unroll_y=0, load_A='BufferToLDS', load_B='BufferToLDS', storeLDS_D=False, betaInFma=True, scheduler='Priority', schedulerCost='LinearWeighted', prefetch=True, prefetchInFlight=4, prefetchLDSFactor=1, prefetchMixMemOps=True, loadScale_A='BufferToVGPR', loadScale_B='BufferToVGPR', swizzleScale=False, swizzleTileSize={'m': 0, 'k': 0, 'n': 0, 'l': 0}, prefetchScale=False, streamK=False, numWGs=0, streamKTwoTile=False, streamKTwoTileDPFirst=False, architecture={'ArchString': 'gfx950', 'Xnack': False, 'Sramecc': True}, matchMemoryAccess=True, version='')
+   0.52% | p=1.1861e-09 
	| 3. FloatsGEMM(M: 4096, N: 4096, K: 32768, alpha: 2, beta: 0, types: {'type_A': 'fp8', 'type_B': 'fp8', 'type_C': 'half', 'type_D': 'half', 'type_acc': 'float', 'trans_A': 'T', 'trans_B': 'N', 'scale_A': 'Separate', 'scaleType_A': 'E8M0', 'scale_B': 'Separate', 'scaleType_B': 'E8M0', 'scaleBlockSize': 32, 'scaleSkipPermlane': False, 'scaleShuffleTileA': [], 'scaleShuffleTileB': []}, scaleValue_A: 1, scaleValue_B: 1, initMode_A: DataInitMode(Bounded), initMode_B: DataInitMode(Bounded), initMode_C: DataInitMode(Bounded), workgroupMappingDim: -1, workgroupMappingValue: -1)| GEMM(mac_m: 128, mac_n: 128, mac_k: 128, wave_m: 16, wave_n: 16, wave_k: 128, wave_b: 1, workgroup_size_x: 128, workgroup_size_y: 2, workgroupRemapXCC: False, workgroupRemapXCCValue: -1, unroll_x: 0, unroll_y: 0, load_A: BufferToLDS, load_B: BufferToLDS, storeLDS_D: False, betaInFma: True, scheduler: Priority, schedulerCost: LinearWeighted, prefetch: True, prefetchInFlight: 4, prefetchLDSFactor: 1, prefetchMixMemOps: True, loadScale_A: BufferToVGPR, loadScale_B: BufferToVGPR, swizzleScale: True, swizzleTileSize: {'m': 64, 'k': 16, 'n': 64, 'l': 16}, prefetchScale: True, streamK: False, numWGs: 0, streamKTwoTile: False, streamKTwoTileDPFirst: False, architecture: {'ArchString': 'gfx950', 'Xnack': False, 'Sramecc': True}, matchMemoryAccess: True, version: ) | GEMMProblem(M=4096, N=4096, K=32768, alpha=2, beta=0, types={'type_A': 'fp8', 'type_B': 'fp8', 'type_C': 'half', 'type_D': 'half', 'type_acc': 'float', 'trans_A': 'T', 'trans_B': 'N', 'scale_A': 'Separate', 'scaleType_A': 'E8M0', 'scale_B': 'Separate', 'scaleType_B': 'E8M0', 'scaleBlockSize': 32, 'scaleSkipPermlane': False, 'scaleShuffleTileA': [], 'scaleShuffleTileB': []}, scaleValue_A=1, scaleValue_B=1, initMode_A='DataInitMode(Bounded)', initMode_B='DataInitMode(Bounded)', initMode_C='DataInitMode(Bounded)', workgroupMappingDim=-1, workgroupMappingValue=-1)GEMMSolution(mac_m=128, mac_n=128, mac_k=128, wave_m=16, wave_n=16, wave_k=128, wave_b=1, workgroup_size_x=128, workgroup_size_y=2, workgroupRemapXCC=False, workgroupRemapXCCValue=-1, unroll_x=0, unroll_y=0, load_A='BufferToLDS', load_B='BufferToLDS', storeLDS_D=False, betaInFma=True, scheduler='Priority', schedulerCost='LinearWeighted', prefetch=True, prefetchInFlight=4, prefetchLDSFactor=1, prefetchMixMemOps=True, loadScale_A='BufferToVGPR', loadScale_B='BufferToVGPR', swizzleScale=True, swizzleTileSize={'m': 64, 'k': 16, 'n': 64, 'l': 16}, prefetchScale=True, streamK=False, numWGs=0, streamKTwoTile=False, streamKTwoTileDPFirst=False, architecture={'ArchString': 'gfx950', 'Xnack': False, 'Sramecc': True}, matchMemoryAccess=True, version='')
+   0.18% | p=1.8651e-26 
	| 3. FloatsGEMM(M: 4096, N: 4096, K: 32768, alpha: 2, beta: 0, types: {'type_A': 'fp8', 'type_B': 'fp8', 'type_C': 'half', 'type_D': 'half', 'type_acc': 'float', 'trans_A': 'T', 'trans_B': 'N', 'scale_A': 'Separate', 'scaleType_A': 'E8M0', 'scale_B': 'Separate', 'scaleType_B': 'E8M0', 'scaleBlockSize': 32, 'scaleSkipPermlane': False, 'scaleShuffleTileA': [], 'scaleShuffleTileB': []}, scaleValue_A: 1, scaleValue_B: 1, initMode_A: DataInitMode(Bounded), initMode_B: DataInitMode(Bounded), initMode_C: DataInitMode(Bounded), workgroupMappingDim: -1, workgroupMappingValue: -1)| GEMM(mac_m: 128, mac_n: 128, mac_k: 128, wave_m: 32, wave_n: 32, wave_k: 64, wave_b: 1, workgroup_size_x: 128, workgroup_size_y: 2, workgroupRemapXCC: False, workgroupRemapXCCValue: -1, unroll_x: 0, unroll_y: 0, load_A: BufferToLDS, load_B: BufferToLDS, storeLDS_D: False, betaInFma: True, scheduler: Priority, schedulerCost: LinearWeighted, prefetch: True, prefetchInFlight: 4, prefetchLDSFactor: 1, prefetchMixMemOps: True, loadScale_A: BufferToVGPR, loadScale_B: BufferToVGPR, swizzleScale: True, swizzleTileSize: {'m': 64, 'k': 16, 'n': 64, 'l': 16}, prefetchScale: True, streamK: False, numWGs: 0, streamKTwoTile: False, streamKTwoTileDPFirst: False, architecture: {'ArchString': 'gfx950', 'Xnack': False, 'Sramecc': True}, matchMemoryAccess: True, version: ) | GEMMProblem(M=4096, N=4096, K=32768, alpha=2, beta=0, types={'type_A': 'fp8', 'type_B': 'fp8', 'type_C': 'half', 'type_D': 'half', 'type_acc': 'float', 'trans_A': 'T', 'trans_B': 'N', 'scale_A': 'Separate', 'scaleType_A': 'E8M0', 'scale_B': 'Separate', 'scaleType_B': 'E8M0', 'scaleBlockSize': 32, 'scaleSkipPermlane': False, 'scaleShuffleTileA': [], 'scaleShuffleTileB': []}, scaleValue_A=1, scaleValue_B=1, initMode_A='DataInitMode(Bounded)', initMode_B='DataInitMode(Bounded)', initMode_C='DataInitMode(Bounded)', workgroupMappingDim=-1, workgroupMappingValue=-1)GEMMSolution(mac_m=128, mac_n=128, mac_k=128, wave_m=32, wave_n=32, wave_k=64, wave_b=1, workgroup_size_x=128, workgroup_size_y=2, workgroupRemapXCC=False, workgroupRemapXCCValue=-1, unroll_x=0, unroll_y=0, load_A='BufferToLDS', load_B='BufferToLDS', storeLDS_D=False, betaInFma=True, scheduler='Priority', schedulerCost='LinearWeighted', prefetch=True, prefetchInFlight=4, prefetchLDSFactor=1, prefetchMixMemOps=True, loadScale_A='BufferToVGPR', loadScale_B='BufferToVGPR', swizzleScale=True, swizzleTileSize={'m': 64, 'k': 16, 'n': 64, 'l': 16}, prefetchScale=True, streamK=False, numWGs=0, streamKTwoTile=False, streamKTwoTileDPFirst=False, architecture={'ArchString': 'gfx950', 'Xnack': False, 'Sramecc': True}, matchMemoryAccess=True, version='')
+   0.23% | p=2.2561e-02 
	| 3. FloatsGEMM(M: 4096, N: 4096, K: 32768, alpha: 2, beta: 0, types: {'type_A': 'fp8', 'type_B': 'fp8', 'type_C': 'half', 'type_D': 'half', 'type_acc': 'float', 'trans_A': 'T', 'trans_B': 'N', 'scale_A': 'Separate', 'scaleType_A': 'E8M0', 'scale_B': 'Separate', 'scaleType_B': 'E8M0', 'scaleBlockSize': 32, 'scaleSkipPermlane': False, 'scaleShuffleTileA': [], 'scaleShuffleTileB': []}, scaleValue_A: 1, scaleValue_B: 1, initMode_A: DataInitMode(Bounded), initMode_B: DataInitMode(Bounded), initMode_C: DataInitMode(Bounded), workgroupMappingDim: 1, workgroupMappingValue: -1)| GEMM(mac_m: 128, mac_n: 128, mac_k: 128, wave_m: 16, wave_n: 16, wave_k: 128, wave_b: 1, workgroup_size_x: 128, workgroup_size_y: 2, workgroupRemapXCC: False, workgroupRemapXCCValue: -1, unroll_x: 0, unroll_y: 0, load_A: BufferToLDS, load_B: BufferToLDS, storeLDS_D: False, betaInFma: True, scheduler: Priority, schedulerCost: LinearWeighted, prefetch: True, prefetchInFlight: 4, prefetchLDSFactor: 1, prefetchMixMemOps: True, loadScale_A: BufferToVGPR, loadScale_B: BufferToVGPR, swizzleScale: True, swizzleTileSize: {'m': 64, 'k': 16, 'n': 64, 'l': 16}, prefetchScale: True, streamK: False, numWGs: 0, streamKTwoTile: False, streamKTwoTileDPFirst: False, architecture: {'ArchString': 'gfx950', 'Xnack': False, 'Sramecc': True}, matchMemoryAccess: True, version: ) | GEMMProblem(M=4096, N=4096, K=32768, alpha=2, beta=0, types={'type_A': 'fp8', 'type_B': 'fp8', 'type_C': 'half', 'type_D': 'half', 'type_acc': 'float', 'trans_A': 'T', 'trans_B': 'N', 'scale_A': 'Separate', 'scaleType_A': 'E8M0', 'scale_B': 'Separate', 'scaleType_B': 'E8M0', 'scaleBlockSize': 32, 'scaleSkipPermlane': False, 'scaleShuffleTileA': [], 'scaleShuffleTileB': []}, scaleValue_A=1, scaleValue_B=1, initMode_A='DataInitMode(Bounded)', initMode_B='DataInitMode(Bounded)', initMode_C='DataInitMode(Bounded)', workgroupMappingDim=1, workgroupMappingValue=-1)GEMMSolution(mac_m=128, mac_n=128, mac_k=128, wave_m=16, wave_n=16, wave_k=128, wave_b=1, workgroup_size_x=128, workgroup_size_y=2, workgroupRemapXCC=False, workgroupRemapXCCValue=-1, unroll_x=0, unroll_y=0, load_A='BufferToLDS', load_B='BufferToLDS', storeLDS_D=False, betaInFma=True, scheduler='Priority', schedulerCost='LinearWeighted', prefetch=True, prefetchInFlight=4, prefetchLDSFactor=1, prefetchMixMemOps=True, loadScale_A='BufferToVGPR', loadScale_B='BufferToVGPR', swizzleScale=True, swizzleTileSize={'m': 64, 'k': 16, 'n': 64, 'l': 16}, prefetchScale=True, streamK=False, numWGs=0, streamKTwoTile=False, streamKTwoTileDPFirst=False, architecture={'ArchString': 'gfx950', 'Xnack': False, 'Sramecc': True}, matchMemoryAccess=True, version='')
+   0.29% | p=1.4393e-04 
	| 3. FloatsGEMM(M: 4096, N: 4096, K: 32768, alpha: 2, beta: 0, types: {'type_A': 'fp8', 'type_B': 'fp8', 'type_C': 'half', 'type_D': 'half', 'type_acc': 'float', 'trans_A': 'T', 'trans_B': 'N', 'scale_A': 'Separate', 'scaleType_A': 'E8M0', 'scale_B': 'Separate', 'scaleType_B': 'E8M0', 'scaleBlockSize': 32, 'scaleSkipPermlane': True, 'scaleShuffleTileA': [64, 4, 4], 'scaleShuffleTileB': [64, 4, 4]}, scaleValue_A: 1, scaleValue_B: 1, initMode_A: DataInitMode(Bounded), initMode_B: DataInitMode(Bounded), initMode_C: DataInitMode(Bounded), workgroupMappingDim: 1, workgroupMappingValue: -1)| GEMM(mac_m: 128, mac_n: 128, mac_k: 128, wave_m: 16, wave_n: 16, wave_k: 128, wave_b: 1, workgroup_size_x: 128, workgroup_size_y: 2, workgroupRemapXCC: False, workgroupRemapXCCValue: -1, unroll_x: 0, unroll_y: 0, load_A: BufferToLDS, load_B: BufferToLDS, storeLDS_D: False, betaInFma: True, scheduler: Priority, schedulerCost: LinearWeighted, prefetch: True, prefetchInFlight: 4, prefetchLDSFactor: 1, prefetchMixMemOps: True, loadScale_A: BufferToVGPR, loadScale_B: BufferToVGPR, swizzleScale: True, swizzleTileSize: {'m': 64, 'k': 16, 'n': 64, 'l': 16}, prefetchScale: True, streamK: False, numWGs: 0, streamKTwoTile: False, streamKTwoTileDPFirst: False, architecture: {'ArchString': 'gfx950', 'Xnack': False, 'Sramecc': True}, matchMemoryAccess: True, version: ) | GEMMProblem(M=4096, N=4096, K=32768, alpha=2, beta=0, types={'type_A': 'fp8', 'type_B': 'fp8', 'type_C': 'half', 'type_D': 'half', 'type_acc': 'float', 'trans_A': 'T', 'trans_B': 'N', 'scale_A': 'Separate', 'scaleType_A': 'E8M0', 'scale_B': 'Separate', 'scaleType_B': 'E8M0', 'scaleBlockSize': 32, 'scaleSkipPermlane': True, 'scaleShuffleTileA': [64, 4, 4], 'scaleShuffleTileB': [64, 4, 4]}, scaleValue_A=1, scaleValue_B=1, initMode_A='DataInitMode(Bounded)', initMode_B='DataInitMode(Bounded)', initMode_C='DataInitMode(Bounded)', workgroupMappingDim=1, workgroupMappingValue=-1)GEMMSolution(mac_m=128, mac_n=128, mac_k=128, wave_m=16, wave_n=16, wave_k=128, wave_b=1, workgroup_size_x=128, workgroup_size_y=2, workgroupRemapXCC=False, workgroupRemapXCCValue=-1, unroll_x=0, unroll_y=0, load_A='BufferToLDS', load_B='BufferToLDS', storeLDS_D=False, betaInFma=True, scheduler='Priority', schedulerCost='LinearWeighted', prefetch=True, prefetchInFlight=4, prefetchLDSFactor=1, prefetchMixMemOps=True, loadScale_A='BufferToVGPR', loadScale_B='BufferToVGPR', swizzleScale=True, swizzleTileSize={'m': 64, 'k': 16, 'n': 64, 'l': 16}, prefetchScale=True, streamK=False, numWGs=0, streamKTwoTile=False, streamKTwoTileDPFirst=False, architecture={'ArchString': 'gfx950', 'Xnack': False, 'Sramecc': True}, matchMemoryAccess=True, version='')
-   0.34% | p=6.2764e-06 
	| 3. FloatsGEMM(M: 4096, N: 4096, K: 32768, alpha: 2, beta: 0.5, types: {'type_A': 'fp4', 'type_B': 'fp4', 'type_C': 'half', 'type_D': 'half', 'type_acc': 'float', 'trans_A': 'T', 'trans_B': 'N', 'scale_A': 'Separate', 'scaleType_A': 'E8M0', 'scale_B': 'Separate', 'scaleType_B': 'E8M0', 'scaleBlockSize': 32, 'scaleSkipPermlane': False, 'scaleShuffleTileA': [], 'scaleShuffleTileB': []}, scaleValue_A: 1, scaleValue_B: 1, initMode_A: DataInitMode(Bounded), initMode_B: DataInitMode(Bounded), initMode_C: DataInitMode(Bounded), workgroupMappingDim: -1, workgroupMappingValue: -1)| GEMM(mac_m: 256, mac_n: 256, mac_k: 128, wave_m: 32, wave_n: 32, wave_k: 64, wave_b: 1, workgroup_size_x: 128, workgroup_size_y: 2, workgroupRemapXCC: False, workgroupRemapXCCValue: -1, unroll_x: 0, unroll_y: 0, load_A: BufferToLDS, load_B: BufferToLDS, storeLDS_D: True, betaInFma: True, scheduler: Priority, schedulerCost: LinearWeighted, prefetch: True, prefetchInFlight: 2, prefetchLDSFactor: 1, prefetchMixMemOps: False, loadScale_A: BufferToLDSViaVGPR, loadScale_B: BufferToLDSViaVGPR, swizzleScale: False, swizzleTileSize: {'m': 64, 'k': 8, 'n': 64, 'l': 8}, prefetchScale: False, streamK: False, numWGs: 0, streamKTwoTile: False, streamKTwoTileDPFirst: False, architecture: {'ArchString': 'gfx950', 'Xnack': False, 'Sramecc': True}, matchMemoryAccess: True, version: ) | GEMMProblem(M=4096, N=4096, K=32768, alpha=2, beta=0.5, types={'type_A': 'fp4', 'type_B': 'fp4', 'type_C': 'half', 'type_D': 'half', 'type_acc': 'float', 'trans_A': 'T', 'trans_B': 'N', 'scale_A': 'Separate', 'scaleType_A': 'E8M0', 'scale_B': 'Separate', 'scaleType_B': 'E8M0', 'scaleBlockSize': 32, 'scaleSkipPermlane': False, 'scaleShuffleTileA': [], 'scaleShuffleTileB': []}, scaleValue_A=1, scaleValue_B=1, initMode_A='DataInitMode(Bounded)', initMode_B='DataInitMode(Bounded)', initMode_C='DataInitMode(Bounded)', workgroupMappingDim=-1, workgroupMappingValue=-1)GEMMSolution(mac_m=256, mac_n=256, mac_k=128, wave_m=32, wave_n=32, wave_k=64, wave_b=1, workgroup_size_x=128, workgroup_size_y=2, workgroupRemapXCC=False, workgroupRemapXCCValue=-1, unroll_x=0, unroll_y=0, load_A='BufferToLDS', load_B='BufferToLDS', storeLDS_D=True, betaInFma=True, scheduler='Priority', schedulerCost='LinearWeighted', prefetch=True, prefetchInFlight=2, prefetchLDSFactor=1, prefetchMixMemOps=False, loadScale_A='BufferToLDSViaVGPR', loadScale_B='BufferToLDSViaVGPR', swizzleScale=False, swizzleTileSize={'m': 64, 'k': 8, 'n': 64, 'l': 8}, prefetchScale=False, streamK=False, numWGs=0, streamKTwoTile=False, streamKTwoTileDPFirst=False, architecture={'ArchString': 'gfx950', 'Xnack': False, 'Sramecc': True}, matchMemoryAccess=True, version='')
-   0.56% | p=1.4473e-14 
	| 3. FloatsGEMM(M: 4096, N: 4096, K: 32768, alpha: 2, beta: 0.5, types: {'type_A': 'fp4', 'type_B': 'fp4', 'type_C': 'half', 'type_D': 'half', 'type_acc': 'float', 'trans_A': 'T', 'trans_B': 'N', 'scale_A': 'Separate', 'scaleType_A': 'E8M0', 'scale_B': 'Separate', 'scaleType_B': 'E8M0', 'scaleBlockSize': 32, 'scaleSkipPermlane': False, 'scaleShuffleTileA': [], 'scaleShuffleTileB': []}, scaleValue_A: 1, scaleValue_B: 1, initMode_A: DataInitMode(Bounded), initMode_B: DataInitMode(Bounded), initMode_C: DataInitMode(Bounded), workgroupMappingDim: -1, workgroupMappingValue: -1)| GEMM(mac_m: 256, mac_n: 256, mac_k: 128, wave_m: 32, wave_n: 32, wave_k: 64, wave_b: 1, workgroup_size_x: 128, workgroup_size_y: 2, workgroupRemapXCC: False, workgroupRemapXCCValue: -1, unroll_x: 0, unroll_y: 0, load_A: BufferToLDSViaVGPR, load_B: BufferToLDSViaVGPR, storeLDS_D: True, betaInFma: True, scheduler: Priority, schedulerCost: LinearWeighted, prefetch: True, prefetchInFlight: 2, prefetchLDSFactor: 2, prefetchMixMemOps: False, loadScale_A: BufferToLDSViaVGPR, loadScale_B: BufferToLDSViaVGPR, swizzleScale: False, swizzleTileSize: {'m': 64, 'k': 8, 'n': 64, 'l': 8}, prefetchScale: False, streamK: False, numWGs: 0, streamKTwoTile: False, streamKTwoTileDPFirst: False, architecture: {'ArchString': 'gfx950', 'Xnack': False, 'Sramecc': True}, matchMemoryAccess: True, version: ) | GEMMProblem(M=4096, N=4096, K=32768, alpha=2, beta=0.5, types={'type_A': 'fp4', 'type_B': 'fp4', 'type_C': 'half', 'type_D': 'half', 'type_acc': 'float', 'trans_A': 'T', 'trans_B': 'N', 'scale_A': 'Separate', 'scaleType_A': 'E8M0', 'scale_B': 'Separate', 'scaleType_B': 'E8M0', 'scaleBlockSize': 32, 'scaleSkipPermlane': False, 'scaleShuffleTileA': [], 'scaleShuffleTileB': []}, scaleValue_A=1, scaleValue_B=1, initMode_A='DataInitMode(Bounded)', initMode_B='DataInitMode(Bounded)', initMode_C='DataInitMode(Bounded)', workgroupMappingDim=-1, workgroupMappingValue=-1)GEMMSolution(mac_m=256, mac_n=256, mac_k=128, wave_m=32, wave_n=32, wave_k=64, wave_b=1, workgroup_size_x=128, workgroup_size_y=2, workgroupRemapXCC=False, workgroupRemapXCCValue=-1, unroll_x=0, unroll_y=0, load_A='BufferToLDSViaVGPR', load_B='BufferToLDSViaVGPR', storeLDS_D=True, betaInFma=True, scheduler='Priority', schedulerCost='LinearWeighted', prefetch=True, prefetchInFlight=2, prefetchLDSFactor=2, prefetchMixMemOps=False, loadScale_A='BufferToLDSViaVGPR', loadScale_B='BufferToLDSViaVGPR', swizzleScale=False, swizzleTileSize={'m': 64, 'k': 8, 'n': 64, 'l': 8}, prefetchScale=False, streamK=False, numWGs=0, streamKTwoTile=False, streamKTwoTileDPFirst=False, architecture={'ArchString': 'gfx950', 'Xnack': False, 'Sramecc': True}, matchMemoryAccess=True, version='')
+   0.20% | p=4.0937e-18 
	| 3. FloatsGEMM(M: 4096, N: 4096, K: 32768, alpha: 2, beta: 0.5, types: {'type_A': 'fp8', 'type_B': 'fp8', 'type_C': 'half', 'type_D': 'half', 'type_acc': 'float', 'trans_A': 'T', 'trans_B': 'N', 'scale_A': 'Separate', 'scaleType_A': 'E8M0', 'scale_B': 'Separate', 'scaleType_B': 'E8M0', 'scaleBlockSize': 32, 'scaleSkipPermlane': False, 'scaleShuffleTileA': [], 'scaleShuffleTileB': []}, scaleValue_A: 1, scaleValue_B: 1, initMode_A: DataInitMode(Bounded), initMode_B: DataInitMode(Bounded), initMode_C: DataInitMode(Bounded), workgroupMappingDim: -1, workgroupMappingValue: -1)| GEMM(mac_m: 128, mac_n: 256, mac_k: 128, wave_m: 16, wave_n: 16, wave_k: 128, wave_b: 1, workgroup_size_x: 128, workgroup_size_y: 2, workgroupRemapXCC: False, workgroupRemapXCCValue: -1, unroll_x: 0, unroll_y: 0, load_A: BufferToLDSViaVGPR, load_B: BufferToLDSViaVGPR, storeLDS_D: True, betaInFma: True, scheduler: Priority, schedulerCost: LinearWeighted, prefetch: True, prefetchInFlight: 2, prefetchLDSFactor: 2, prefetchMixMemOps: False, loadScale_A: BufferToLDSViaVGPR, loadScale_B: BufferToLDSViaVGPR, swizzleScale: False, swizzleTileSize: {'m': 64, 'k': 8, 'n': 64, 'l': 8}, prefetchScale: False, streamK: False, numWGs: 0, streamKTwoTile: False, streamKTwoTileDPFirst: False, architecture: {'ArchString': 'gfx950', 'Xnack': False, 'Sramecc': True}, matchMemoryAccess: True, version: ) | GEMMProblem(M=4096, N=4096, K=32768, alpha=2, beta=0.5, types={'type_A': 'fp8', 'type_B': 'fp8', 'type_C': 'half', 'type_D': 'half', 'type_acc': 'float', 'trans_A': 'T', 'trans_B': 'N', 'scale_A': 'Separate', 'scaleType_A': 'E8M0', 'scale_B': 'Separate', 'scaleType_B': 'E8M0', 'scaleBlockSize': 32, 'scaleSkipPermlane': False, 'scaleShuffleTileA': [], 'scaleShuffleTileB': []}, scaleValue_A=1, scaleValue_B=1, initMode_A='DataInitMode(Bounded)', initMode_B='DataInitMode(Bounded)', initMode_C='DataInitMode(Bounded)', workgroupMappingDim=-1, workgroupMappingValue=-1)GEMMSolution(mac_m=128, mac_n=256, mac_k=128, wave_m=16, wave_n=16, wave_k=128, wave_b=1, workgroup_size_x=128, workgroup_size_y=2, workgroupRemapXCC=False, workgroupRemapXCCValue=-1, unroll_x=0, unroll_y=0, load_A='BufferToLDSViaVGPR', load_B='BufferToLDSViaVGPR', storeLDS_D=True, betaInFma=True, scheduler='Priority', schedulerCost='LinearWeighted', prefetch=True, prefetchInFlight=2, prefetchLDSFactor=2, prefetchMixMemOps=False, loadScale_A='BufferToLDSViaVGPR', loadScale_B='BufferToLDSViaVGPR', swizzleScale=False, swizzleTileSize={'m': 64, 'k': 8, 'n': 64, 'l': 8}, prefetchScale=False, streamK=False, numWGs=0, streamKTwoTile=False, streamKTwoTileDPFirst=False, architecture={'ArchString': 'gfx950', 'Xnack': False, 'Sramecc': True}, matchMemoryAccess=True, version='')
+   0.28% | p=1.8450e-37 
	| 3. FloatsGEMM(M: 4096, N: 4096, K: 32768, alpha: 2, beta: 0.5, types: {'type_A': 'fp8', 'type_B': 'fp8', 'type_C': 'half', 'type_D': 'half', 'type_acc': 'float', 'trans_A': 'T', 'trans_B': 'N', 'scale_A': 'Separate', 'scaleType_A': 'E8M0', 'scale_B': 'Separate', 'scaleType_B': 'E8M0', 'scaleBlockSize': 32, 'scaleSkipPermlane': False, 'scaleShuffleTileA': [], 'scaleShuffleTileB': []}, scaleValue_A: 1, scaleValue_B: 1, initMode_A: DataInitMode(Bounded), initMode_B: DataInitMode(Bounded), initMode_C: DataInitMode(Bounded), workgroupMappingDim: -1, workgroupMappingValue: -1)| GEMM(mac_m: 256, mac_n: 256, mac_k: 128, wave_m: 32, wave_n: 32, wave_k: 64, wave_b: 1, workgroup_size_x: 128, workgroup_size_y: 2, workgroupRemapXCC: False, workgroupRemapXCCValue: -1, unroll_x: 0, unroll_y: 0, load_A: BufferToLDS, load_B: BufferToLDS, storeLDS_D: True, betaInFma: True, scheduler: Priority, schedulerCost: LinearWeighted, prefetch: True, prefetchInFlight: 2, prefetchLDSFactor: 1, prefetchMixMemOps: False, loadScale_A: BufferToLDSViaVGPR, loadScale_B: BufferToLDSViaVGPR, swizzleScale: False, swizzleTileSize: {'m': 64, 'k': 8, 'n': 64, 'l': 8}, prefetchScale: False, streamK: False, numWGs: 0, streamKTwoTile: False, streamKTwoTileDPFirst: False, architecture: {'ArchString': 'gfx950', 'Xnack': False, 'Sramecc': True}, matchMemoryAccess: True, version: ) | GEMMProblem(M=4096, N=4096, K=32768, alpha=2, beta=0.5, types={'type_A': 'fp8', 'type_B': 'fp8', 'type_C': 'half', 'type_D': 'half', 'type_acc': 'float', 'trans_A': 'T', 'trans_B': 'N', 'scale_A': 'Separate', 'scaleType_A': 'E8M0', 'scale_B': 'Separate', 'scaleType_B': 'E8M0', 'scaleBlockSize': 32, 'scaleSkipPermlane': False, 'scaleShuffleTileA': [], 'scaleShuffleTileB': []}, scaleValue_A=1, scaleValue_B=1, initMode_A='DataInitMode(Bounded)', initMode_B='DataInitMode(Bounded)', initMode_C='DataInitMode(Bounded)', workgroupMappingDim=-1, workgroupMappingValue=-1)GEMMSolution(mac_m=256, mac_n=256, mac_k=128, wave_m=32, wave_n=32, wave_k=64, wave_b=1, workgroup_size_x=128, workgroup_size_y=2, workgroupRemapXCC=False, workgroupRemapXCCValue=-1, unroll_x=0, unroll_y=0, load_A='BufferToLDS', load_B='BufferToLDS', storeLDS_D=True, betaInFma=True, scheduler='Priority', schedulerCost='LinearWeighted', prefetch=True, prefetchInFlight=2, prefetchLDSFactor=1, prefetchMixMemOps=False, loadScale_A='BufferToLDSViaVGPR', loadScale_B='BufferToLDSViaVGPR', swizzleScale=False, swizzleTileSize={'m': 64, 'k': 8, 'n': 64, 'l': 8}, prefetchScale=False, streamK=False, numWGs=0, streamKTwoTile=False, streamKTwoTileDPFirst=False, architecture={'ArchString': 'gfx950', 'Xnack': False, 'Sramecc': True}, matchMemoryAccess=True, version='')
+   0.08% | p=1.2002e-04 
	| 3. FloatsGEMM(M: 4096, N: 4096, K: 32768, alpha: 2, beta: 0.5, types: {'type_A': 'fp8', 'type_B': 'fp8', 'type_C': 'half', 'type_D': 'half', 'type_acc': 'float', 'trans_A': 'T', 'trans_B': 'N', 'scale_A': 'Separate', 'scaleType_A': 'E8M0', 'scale_B': 'Separate', 'scaleType_B': 'E8M0', 'scaleBlockSize': 32, 'scaleSkipPermlane': False, 'scaleShuffleTileA': [], 'scaleShuffleTileB': []}, scaleValue_A: 1, scaleValue_B: 1, initMode_A: DataInitMode(Bounded), initMode_B: DataInitMode(Bounded), initMode_C: DataInitMode(Bounded), workgroupMappingDim: 0, workgroupMappingValue: -1)| GEMM(mac_m: 256, mac_n: 256, mac_k: 128, wave_m: 32, wave_n: 32, wave_k: 64, wave_b: 1, workgroup_size_x: 128, workgroup_size_y: 2, workgroupRemapXCC: False, workgroupRemapXCCValue: -1, unroll_x: 0, unroll_y: 0, load_A: BufferToLDS, load_B: BufferToLDS, storeLDS_D: True, betaInFma: True, scheduler: Priority, schedulerCost: LinearWeighted, prefetch: True, prefetchInFlight: 2, prefetchLDSFactor: 1, prefetchMixMemOps: False, loadScale_A: BufferToLDSViaVGPR, loadScale_B: BufferToLDSViaVGPR, swizzleScale: False, swizzleTileSize: {'m': 64, 'k': 8, 'n': 64, 'l': 8}, prefetchScale: False, streamK: False, numWGs: 0, streamKTwoTile: False, streamKTwoTileDPFirst: False, architecture: {'ArchString': 'gfx950', 'Xnack': False, 'Sramecc': True}, matchMemoryAccess: True, version: ) | GEMMProblem(M=4096, N=4096, K=32768, alpha=2, beta=0.5, types={'type_A': 'fp8', 'type_B': 'fp8', 'type_C': 'half', 'type_D': 'half', 'type_acc': 'float', 'trans_A': 'T', 'trans_B': 'N', 'scale_A': 'Separate', 'scaleType_A': 'E8M0', 'scale_B': 'Separate', 'scaleType_B': 'E8M0', 'scaleBlockSize': 32, 'scaleSkipPermlane': False, 'scaleShuffleTileA': [], 'scaleShuffleTileB': []}, scaleValue_A=1, scaleValue_B=1, initMode_A='DataInitMode(Bounded)', initMode_B='DataInitMode(Bounded)', initMode_C='DataInitMode(Bounded)', workgroupMappingDim=0, workgroupMappingValue=-1)GEMMSolution(mac_m=256, mac_n=256, mac_k=128, wave_m=32, wave_n=32, wave_k=64, wave_b=1, workgroup_size_x=128, workgroup_size_y=2, workgroupRemapXCC=False, workgroupRemapXCCValue=-1, unroll_x=0, unroll_y=0, load_A='BufferToLDS', load_B='BufferToLDS', storeLDS_D=True, betaInFma=True, scheduler='Priority', schedulerCost='LinearWeighted', prefetch=True, prefetchInFlight=2, prefetchLDSFactor=1, prefetchMixMemOps=False, loadScale_A='BufferToLDSViaVGPR', loadScale_B='BufferToLDSViaVGPR', swizzleScale=False, swizzleTileSize={'m': 64, 'k': 8, 'n': 64, 'l': 8}, prefetchScale=False, streamK=False, numWGs=0, streamKTwoTile=False, streamKTwoTileDPFirst=False, architecture={'ArchString': 'gfx950', 'Xnack': False, 'Sramecc': True}, matchMemoryAccess=True, version='')
+   0.27% | p=1.7451e-03 
	| 3. FloatsGEMM(M: 4096, N: 4096, K: 8192, alpha: 2, beta: 0.5, types: {'type_A': 'bf6', 'type_B': 'bf6', 'type_C': 'float', 'type_D': 'float', 'type_acc': 'float', 'trans_A': 'N', 'trans_B': 'T', 'scale_A': 'None', 'scaleType_A': 'None', 'scale_B': 'None', 'scaleType_B': 'None', 'scaleBlockSize': -1, 'scaleSkipPermlane': False, 'scaleShuffleTileA': [], 'scaleShuffleTileB': []}, scaleValue_A: 1, scaleValue_B: 1, initMode_A: DataInitMode(Bounded), initMode_B: DataInitMode(Bounded), initMode_C: DataInitMode(Bounded), workgroupMappingDim: -1, workgroupMappingValue: -1)| GEMM(mac_m: 128, mac_n: 128, mac_k: 64, wave_m: 32, wave_n: 32, wave_k: 64, wave_b: 1, workgroup_size_x: 256, workgroup_size_y: 1, workgroupRemapXCC: False, workgroupRemapXCCValue: -1, unroll_x: 0, unroll_y: 0, load_A: BufferToLDSViaVGPR, load_B: BufferToLDSViaVGPR, storeLDS_D: True, betaInFma: True, scheduler: Priority, schedulerCost: LinearWeighted, prefetch: True, prefetchInFlight: 2, prefetchLDSFactor: 0, prefetchMixMemOps: False, loadScale_A: BufferToVGPR, loadScale_B: BufferToVGPR, swizzleScale: False, swizzleTileSize: {'m': 0, 'k': 0, 'n': 0, 'l': 0}, prefetchScale: False, streamK: False, numWGs: 0, streamKTwoTile: False, streamKTwoTileDPFirst: False, architecture: {'ArchString': 'gfx950', 'Xnack': False, 'Sramecc': True}, matchMemoryAccess: True, version: ) | GEMMProblem(M=4096, N=4096, K=8192, alpha=2, beta=0.5, types={'type_A': 'bf6', 'type_B': 'bf6', 'type_C': 'float', 'type_D': 'float', 'type_acc': 'float', 'trans_A': 'N', 'trans_B': 'T', 'scale_A': 'None', 'scaleType_A': 'None', 'scale_B': 'None', 'scaleType_B': 'None', 'scaleBlockSize': -1, 'scaleSkipPermlane': False, 'scaleShuffleTileA': [], 'scaleShuffleTileB': []}, scaleValue_A=1, scaleValue_B=1, initMode_A='DataInitMode(Bounded)', initMode_B='DataInitMode(Bounded)', initMode_C='DataInitMode(Bounded)', workgroupMappingDim=-1, workgroupMappingValue=-1)GEMMSolution(mac_m=128, mac_n=128, mac_k=64, wave_m=32, wave_n=32, wave_k=64, wave_b=1, workgroup_size_x=256, workgroup_size_y=1, workgroupRemapXCC=False, workgroupRemapXCCValue=-1, unroll_x=0, unroll_y=0, load_A='BufferToLDSViaVGPR', load_B='BufferToLDSViaVGPR', storeLDS_D=True, betaInFma=True, scheduler='Priority', schedulerCost='LinearWeighted', prefetch=True, prefetchInFlight=2, prefetchLDSFactor=0, prefetchMixMemOps=False, loadScale_A='BufferToVGPR', loadScale_B='BufferToVGPR', swizzleScale=False, swizzleTileSize={'m': 0, 'k': 0, 'n': 0, 'l': 0}, prefetchScale=False, streamK=False, numWGs=0, streamKTwoTile=False, streamKTwoTileDPFirst=False, architecture={'ArchString': 'gfx950', 'Xnack': False, 'Sramecc': True}, matchMemoryAccess=True, version='')
+   1.40% | p=1.7451e-03 
	| 3. FloatsGEMM(M: 4096, N: 4096, K: 8192, alpha: 2, beta: 0.5, types: {'type_A': 'bf6', 'type_B': 'bf6', 'type_C': 'float', 'type_D': 'float', 'type_acc': 'float', 'trans_A': 'T', 'trans_B': 'N', 'scale_A': 'None', 'scaleType_A': 'None', 'scale_B': 'None', 'scaleType_B': 'None', 'scaleBlockSize': -1, 'scaleSkipPermlane': False, 'scaleShuffleTileA': [], 'scaleShuffleTileB': []}, scaleValue_A: 1, scaleValue_B: 1, initMode_A: DataInitMode(Bounded), initMode_B: DataInitMode(Bounded), initMode_C: DataInitMode(Bounded), workgroupMappingDim: -1, workgroupMappingValue: -1)| GEMM(mac_m: 128, mac_n: 128, mac_k: 64, wave_m: 32, wave_n: 32, wave_k: 64, wave_b: 1, workgroup_size_x: 256, workgroup_size_y: 1, workgroupRemapXCC: False, workgroupRemapXCCValue: -1, unroll_x: 0, unroll_y: 0, load_A: BufferToLDSViaVGPR, load_B: BufferToLDSViaVGPR, storeLDS_D: True, betaInFma: True, scheduler: Priority, schedulerCost: LinearWeighted, prefetch: True, prefetchInFlight: 2, prefetchLDSFactor: 0, prefetchMixMemOps: False, loadScale_A: BufferToVGPR, loadScale_B: BufferToVGPR, swizzleScale: False, swizzleTileSize: {'m': 0, 'k': 0, 'n': 0, 'l': 0}, prefetchScale: False, streamK: False, numWGs: 0, streamKTwoTile: False, streamKTwoTileDPFirst: False, architecture: {'ArchString': 'gfx950', 'Xnack': False, 'Sramecc': True}, matchMemoryAccess: True, version: ) | GEMMProblem(M=4096, N=4096, K=8192, alpha=2, beta=0.5, types={'type_A': 'bf6', 'type_B': 'bf6', 'type_C': 'float', 'type_D': 'float', 'type_acc': 'float', 'trans_A': 'T', 'trans_B': 'N', 'scale_A': 'None', 'scaleType_A': 'None', 'scale_B': 'None', 'scaleType_B': 'None', 'scaleBlockSize': -1, 'scaleSkipPermlane': False, 'scaleShuffleTileA': [], 'scaleShuffleTileB': []}, scaleValue_A=1, scaleValue_B=1, initMode_A='DataInitMode(Bounded)', initMode_B='DataInitMode(Bounded)', initMode_C='DataInitMode(Bounded)', workgroupMappingDim=-1, workgroupMappingValue=-1)GEMMSolution(mac_m=128, mac_n=128, mac_k=64, wave_m=32, wave_n=32, wave_k=64, wave_b=1, workgroup_size_x=256, workgroup_size_y=1, workgroupRemapXCC=False, workgroupRemapXCCValue=-1, unroll_x=0, unroll_y=0, load_A='BufferToLDSViaVGPR', load_B='BufferToLDSViaVGPR', storeLDS_D=True, betaInFma=True, scheduler='Priority', schedulerCost='LinearWeighted', prefetch=True, prefetchInFlight=2, prefetchLDSFactor=0, prefetchMixMemOps=False, loadScale_A='BufferToVGPR', loadScale_B='BufferToVGPR', swizzleScale=False, swizzleTileSize={'m': 0, 'k': 0, 'n': 0, 'l': 0}, prefetchScale=False, streamK=False, numWGs=0, streamKTwoTile=False, streamKTwoTileDPFirst=False, architecture={'ArchString': 'gfx950', 'Xnack': False, 'Sramecc': True}, matchMemoryAccess=True, version='')
-   1.57% | p=1.7451e-03 
	| 3. FloatsGEMM(M: 4096, N: 4096, K: 8192, alpha: 2, beta: 0.5, types: {'type_A': 'fp4', 'type_B': 'fp4', 'type_C': 'float', 'type_D': 'float', 'type_acc': 'float', 'trans_A': 'N', 'trans_B': 'T', 'scale_A': 'None', 'scaleType_A': 'None', 'scale_B': 'None', 'scaleType_B': 'None', 'scaleBlockSize': -1, 'scaleSkipPermlane': False, 'scaleShuffleTileA': [], 'scaleShuffleTileB': []}, scaleValue_A: 1, scaleValue_B: 1, initMode_A: DataInitMode(Bounded), initMode_B: DataInitMode(Bounded), initMode_C: DataInitMode(Bounded), workgroupMappingDim: -1, workgroupMappingValue: -1)| GEMM(mac_m: 128, mac_n: 128, mac_k: 64, wave_m: 32, wave_n: 32, wave_k: 64, wave_b: 1, workgroup_size_x: 256, workgroup_size_y: 1, workgroupRemapXCC: False, workgroupRemapXCCValue: -1, unroll_x: 0, unroll_y: 0, load_A: BufferToLDSViaVGPR, load_B: BufferToLDSViaVGPR, storeLDS_D: True, betaInFma: True, scheduler: Priority, schedulerCost: LinearWeighted, prefetch: True, prefetchInFlight: 2, prefetchLDSFactor: 0, prefetchMixMemOps: False, loadScale_A: BufferToVGPR, loadScale_B: BufferToVGPR, swizzleScale: False, swizzleTileSize: {'m': 0, 'k': 0, 'n': 0, 'l': 0}, prefetchScale: False, streamK: False, numWGs: 0, streamKTwoTile: False, streamKTwoTileDPFirst: False, architecture: {'ArchString': 'gfx950', 'Xnack': False, 'Sramecc': True}, matchMemoryAccess: True, version: ) | GEMMProblem(M=4096, N=4096, K=8192, alpha=2, beta=0.5, types={'type_A': 'fp4', 'type_B': 'fp4', 'type_C': 'float', 'type_D': 'float', 'type_acc': 'float', 'trans_A': 'N', 'trans_B': 'T', 'scale_A': 'None', 'scaleType_A': 'None', 'scale_B': 'None', 'scaleType_B': 'None', 'scaleBlockSize': -1, 'scaleSkipPermlane': False, 'scaleShuffleTileA': [], 'scaleShuffleTileB': []}, scaleValue_A=1, scaleValue_B=1, initMode_A='DataInitMode(Bounded)', initMode_B='DataInitMode(Bounded)', initMode_C='DataInitMode(Bounded)', workgroupMappingDim=-1, workgroupMappingValue=-1)GEMMSolution(mac_m=128, mac_n=128, mac_k=64, wave_m=32, wave_n=32, wave_k=64, wave_b=1, workgroup_size_x=256, workgroup_size_y=1, workgroupRemapXCC=False, workgroupRemapXCCValue=-1, unroll_x=0, unroll_y=0, load_A='BufferToLDSViaVGPR', load_B='BufferToLDSViaVGPR', storeLDS_D=True, betaInFma=True, scheduler='Priority', schedulerCost='LinearWeighted', prefetch=True, prefetchInFlight=2, prefetchLDSFactor=0, prefetchMixMemOps=False, loadScale_A='BufferToVGPR', loadScale_B='BufferToVGPR', swizzleScale=False, swizzleTileSize={'m': 0, 'k': 0, 'n': 0, 'l': 0}, prefetchScale=False, streamK=False, numWGs=0, streamKTwoTile=False, streamKTwoTileDPFirst=False, architecture={'ArchString': 'gfx950', 'Xnack': False, 'Sramecc': True}, matchMemoryAccess=True, version='')
+   1.88% | p=2.5347e-02 
	| 3. FloatsGEMM(M: 4096, N: 4096, K: 8192, alpha: 2, beta: 0.5, types: {'type_A': 'fp4', 'type_B': 'fp4', 'type_C': 'float', 'type_D': 'float', 'type_acc': 'float', 'trans_A': 'T', 'trans_B': 'N', 'scale_A': 'None', 'scaleType_A': 'None', 'scale_B': 'None', 'scaleType_B': 'None', 'scaleBlockSize': -1, 'scaleSkipPermlane': False, 'scaleShuffleTileA': [], 'scaleShuffleTileB': []}, scaleValue_A: 1, scaleValue_B: 1, initMode_A: DataInitMode(Bounded), initMode_B: DataInitMode(Bounded), initMode_C: DataInitMode(Bounded), workgroupMappingDim: -1, workgroupMappingValue: -1)| GEMM(mac_m: 128, mac_n: 128, mac_k: 64, wave_m: 32, wave_n: 32, wave_k: 64, wave_b: 1, workgroup_size_x: 256, workgroup_size_y: 1, workgroupRemapXCC: False, workgroupRemapXCCValue: -1, unroll_x: 0, unroll_y: 0, load_A: BufferToLDSViaVGPR, load_B: BufferToLDSViaVGPR, storeLDS_D: True, betaInFma: True, scheduler: Priority, schedulerCost: LinearWeighted, prefetch: True, prefetchInFlight: 2, prefetchLDSFactor: 0, prefetchMixMemOps: False, loadScale_A: BufferToVGPR, loadScale_B: BufferToVGPR, swizzleScale: False, swizzleTileSize: {'m': 0, 'k': 0, 'n': 0, 'l': 0}, prefetchScale: False, streamK: False, numWGs: 0, streamKTwoTile: False, streamKTwoTileDPFirst: False, architecture: {'ArchString': 'gfx950', 'Xnack': False, 'Sramecc': True}, matchMemoryAccess: True, version: ) | GEMMProblem(M=4096, N=4096, K=8192, alpha=2, beta=0.5, types={'type_A': 'fp4', 'type_B': 'fp4', 'type_C': 'float', 'type_D': 'float', 'type_acc': 'float', 'trans_A': 'T', 'trans_B': 'N', 'scale_A': 'None', 'scaleType_A': 'None', 'scale_B': 'None', 'scaleType_B': 'None', 'scaleBlockSize': -1, 'scaleSkipPermlane': False, 'scaleShuffleTileA': [], 'scaleShuffleTileB': []}, scaleValue_A=1, scaleValue_B=1, initMode_A='DataInitMode(Bounded)', initMode_B='DataInitMode(Bounded)', initMode_C='DataInitMode(Bounded)', workgroupMappingDim=-1, workgroupMappingValue=-1)GEMMSolution(mac_m=128, mac_n=128, mac_k=64, wave_m=32, wave_n=32, wave_k=64, wave_b=1, workgroup_size_x=256, workgroup_size_y=1, workgroupRemapXCC=False, workgroupRemapXCCValue=-1, unroll_x=0, unroll_y=0, load_A='BufferToLDSViaVGPR', load_B='BufferToLDSViaVGPR', storeLDS_D=True, betaInFma=True, scheduler='Priority', schedulerCost='LinearWeighted', prefetch=True, prefetchInFlight=2, prefetchLDSFactor=0, prefetchMixMemOps=False, loadScale_A='BufferToVGPR', loadScale_B='BufferToVGPR', swizzleScale=False, swizzleTileSize={'m': 0, 'k': 0, 'n': 0, 'l': 0}, prefetchScale=False, streamK=False, numWGs=0, streamKTwoTile=False, streamKTwoTileDPFirst=False, architecture={'ArchString': 'gfx950', 'Xnack': False, 'Sramecc': True}, matchMemoryAccess=True, version='')
-   0.42% | p=2.5347e-02 
	| 3. FloatsGEMM(M: 4096, N: 4096, K: 8192, alpha: 2, beta: 0.5, types: {'type_A': 'fp6', 'type_B': 'fp6', 'type_C': 'float', 'type_D': 'float', 'type_acc': 'float', 'trans_A': 'N', 'trans_B': 'N', 'scale_A': 'None', 'scaleType_A': 'None', 'scale_B': 'None', 'scaleType_B': 'None', 'scaleBlockSize': -1, 'scaleSkipPermlane': False, 'scaleShuffleTileA': [], 'scaleShuffleTileB': []}, scaleValue_A: 1, scaleValue_B: 1, initMode_A: DataInitMode(Bounded), initMode_B: DataInitMode(Bounded), initMode_C: DataInitMode(Bounded), workgroupMappingDim: -1, workgroupMappingValue: -1)| GEMM(mac_m: 128, mac_n: 128, mac_k: 64, wave_m: 32, wave_n: 32, wave_k: 64, wave_b: 1, workgroup_size_x: 256, workgroup_size_y: 1, workgroupRemapXCC: False, workgroupRemapXCCValue: -1, unroll_x: 0, unroll_y: 0, load_A: BufferToLDSViaVGPR, load_B: BufferToLDSViaVGPR, storeLDS_D: True, betaInFma: True, scheduler: Priority, schedulerCost: LinearWeighted, prefetch: True, prefetchInFlight: 2, prefetchLDSFactor: 0, prefetchMixMemOps: False, loadScale_A: BufferToVGPR, loadScale_B: BufferToVGPR, swizzleScale: False, swizzleTileSize: {'m': 0, 'k': 0, 'n': 0, 'l': 0}, prefetchScale: False, streamK: False, numWGs: 0, streamKTwoTile: False, streamKTwoTileDPFirst: False, architecture: {'ArchString': 'gfx950', 'Xnack': False, 'Sramecc': True}, matchMemoryAccess: True, version: ) | GEMMProblem(M=4096, N=4096, K=8192, alpha=2, beta=0.5, types={'type_A': 'fp6', 'type_B': 'fp6', 'type_C': 'float', 'type_D': 'float', 'type_acc': 'float', 'trans_A': 'N', 'trans_B': 'N', 'scale_A': 'None', 'scaleType_A': 'None', 'scale_B': 'None', 'scaleType_B': 'None', 'scaleBlockSize': -1, 'scaleSkipPermlane': False, 'scaleShuffleTileA': [], 'scaleShuffleTileB': []}, scaleValue_A=1, scaleValue_B=1, initMode_A='DataInitMode(Bounded)', initMode_B='DataInitMode(Bounded)', initMode_C='DataInitMode(Bounded)', workgroupMappingDim=-1, workgroupMappingValue=-1)GEMMSolution(mac_m=128, mac_n=128, mac_k=64, wave_m=32, wave_n=32, wave_k=64, wave_b=1, workgroup_size_x=256, workgroup_size_y=1, workgroupRemapXCC=False, workgroupRemapXCCValue=-1, unroll_x=0, unroll_y=0, load_A='BufferToLDSViaVGPR', load_B='BufferToLDSViaVGPR', storeLDS_D=True, betaInFma=True, scheduler='Priority', schedulerCost='LinearWeighted', prefetch=True, prefetchInFlight=2, prefetchLDSFactor=0, prefetchMixMemOps=False, loadScale_A='BufferToVGPR', loadScale_B='BufferToVGPR', swizzleScale=False, swizzleTileSize={'m': 0, 'k': 0, 'n': 0, 'l': 0}, prefetchScale=False, streamK=False, numWGs=0, streamKTwoTile=False, streamKTwoTileDPFirst=False, architecture={'ArchString': 'gfx950', 'Xnack': False, 'Sramecc': True}, matchMemoryAccess=True, version='')
+   1.16% | p=2.5347e-02 
	| 3. FloatsGEMM(M: 4096, N: 4096, K: 8192, alpha: 2, beta: 0.5, types: {'type_A': 'fp6', 'type_B': 'fp6', 'type_C': 'float', 'type_D': 'float', 'type_acc': 'float', 'trans_A': 'T', 'trans_B': 'T', 'scale_A': 'None', 'scaleType_A': 'None', 'scale_B': 'None', 'scaleType_B': 'None', 'scaleBlockSize': -1, 'scaleSkipPermlane': False, 'scaleShuffleTileA': [], 'scaleShuffleTileB': []}, scaleValue_A: 1, scaleValue_B: 1, initMode_A: DataInitMode(Bounded), initMode_B: DataInitMode(Bounded), initMode_C: DataInitMode(Bounded), workgroupMappingDim: -1, workgroupMappingValue: -1)| GEMM(mac_m: 64, mac_n: 64, mac_k: 128, wave_m: 16, wave_n: 16, wave_k: 128, wave_b: 1, workgroup_size_x: 256, workgroup_size_y: 1, workgroupRemapXCC: False, workgroupRemapXCCValue: -1, unroll_x: 0, unroll_y: 0, load_A: BufferToLDSViaVGPR, load_B: BufferToLDSViaVGPR, storeLDS_D: True, betaInFma: True, scheduler: Priority, schedulerCost: LinearWeighted, prefetch: True, prefetchInFlight: 2, prefetchLDSFactor: 0, prefetchMixMemOps: False, loadScale_A: BufferToVGPR, loadScale_B: BufferToVGPR, swizzleScale: False, swizzleTileSize: {'m': 0, 'k': 0, 'n': 0, 'l': 0}, prefetchScale: False, streamK: False, numWGs: 0, streamKTwoTile: False, streamKTwoTileDPFirst: False, architecture: {'ArchString': 'gfx950', 'Xnack': False, 'Sramecc': True}, matchMemoryAccess: True, version: ) | GEMMProblem(M=4096, N=4096, K=8192, alpha=2, beta=0.5, types={'type_A': 'fp6', 'type_B': 'fp6', 'type_C': 'float', 'type_D': 'float', 'type_acc': 'float', 'trans_A': 'T', 'trans_B': 'T', 'scale_A': 'None', 'scaleType_A': 'None', 'scale_B': 'None', 'scaleType_B': 'None', 'scaleBlockSize': -1, 'scaleSkipPermlane': False, 'scaleShuffleTileA': [], 'scaleShuffleTileB': []}, scaleValue_A=1, scaleValue_B=1, initMode_A='DataInitMode(Bounded)', initMode_B='DataInitMode(Bounded)', initMode_C='DataInitMode(Bounded)', workgroupMappingDim=-1, workgroupMappingValue=-1)GEMMSolution(mac_m=64, mac_n=64, mac_k=128, wave_m=16, wave_n=16, wave_k=128, wave_b=1, workgroup_size_x=256, workgroup_size_y=1, workgroupRemapXCC=False, workgroupRemapXCCValue=-1, unroll_x=0, unroll_y=0, load_A='BufferToLDSViaVGPR', load_B='BufferToLDSViaVGPR', storeLDS_D=True, betaInFma=True, scheduler='Priority', schedulerCost='LinearWeighted', prefetch=True, prefetchInFlight=2, prefetchLDSFactor=0, prefetchMixMemOps=False, loadScale_A='BufferToVGPR', loadScale_B='BufferToVGPR', swizzleScale=False, swizzleTileSize={'m': 0, 'k': 0, 'n': 0, 'l': 0}, prefetchScale=False, streamK=False, numWGs=0, streamKTwoTile=False, streamKTwoTileDPFirst=False, architecture={'ArchString': 'gfx950', 'Xnack': False, 'Sramecc': True}, matchMemoryAccess=True, version='')
-   0.36% | p=2.5347e-02 
	| 3. FloatsGEMM(M: 7680, N: 8448, K: 8192, alpha: 2, beta: 0.5, types: {'type_A': 'half', 'type_B': 'half', 'type_C': 'half', 'type_D': 'half', 'type_acc': 'float', 'trans_A': 'N', 'trans_B': 'T', 'scale_A': 'None', 'scaleType_A': 'None', 'scale_B': 'None', 'scaleType_B': 'None', 'scaleBlockSize': -1, 'scaleSkipPermlane': False, 'scaleShuffleTileA': [], 'scaleShuffleTileB': []}, scaleValue_A: 1, scaleValue_B: 1, initMode_A: DataInitMode(Bounded), initMode_B: DataInitMode(Bounded), initMode_C: DataInitMode(Bounded), workgroupMappingDim: -1, workgroupMappingValue: -1)| GEMM(mac_m: 128, mac_n: 256, mac_k: 16, wave_m: 32, wave_n: 32, wave_k: 8, wave_b: 1, workgroup_size_x: 128, workgroup_size_y: 2, workgroupRemapXCC: False, workgroupRemapXCCValue: -1, unroll_x: 0, unroll_y: 0, load_A: BufferToLDSViaVGPR, load_B: BufferToLDSViaVGPR, storeLDS_D: True, betaInFma: True, scheduler: Priority, schedulerCost: LinearWeighted, prefetch: False, prefetchInFlight: 2, prefetchLDSFactor: 0, prefetchMixMemOps: False, loadScale_A: BufferToVGPR, loadScale_B: BufferToVGPR, swizzleScale: False, swizzleTileSize: {'m': 0, 'k': 0, 'n': 0, 'l': 0}, prefetchScale: False, streamK: True, numWGs: 256, streamKTwoTile: False, streamKTwoTileDPFirst: False, architecture: {'ArchString': 'gfx950', 'Xnack': False, 'Sramecc': True}, matchMemoryAccess: True, version: ) | GEMMProblem(M=7680, N=8448, K=8192, alpha=2, beta=0.5, types={'type_A': 'half', 'type_B': 'half', 'type_C': 'half', 'type_D': 'half', 'type_acc': 'float', 'trans_A': 'N', 'trans_B': 'T', 'scale_A': 'None', 'scaleType_A': 'None', 'scale_B': 'None', 'scaleType_B': 'None', 'scaleBlockSize': -1, 'scaleSkipPermlane': False, 'scaleShuffleTileA': [], 'scaleShuffleTileB': []}, scaleValue_A=1, scaleValue_B=1, initMode_A='DataInitMode(Bounded)', initMode_B='DataInitMode(Bounded)', initMode_C='DataInitMode(Bounded)', workgroupMappingDim=-1, workgroupMappingValue=-1)GEMMSolution(mac_m=128, mac_n=256, mac_k=16, wave_m=32, wave_n=32, wave_k=8, wave_b=1, workgroup_size_x=128, workgroup_size_y=2, workgroupRemapXCC=False, workgroupRemapXCCValue=-1, unroll_x=0, unroll_y=0, load_A='BufferToLDSViaVGPR', load_B='BufferToLDSViaVGPR', storeLDS_D=True, betaInFma=True, scheduler='Priority', schedulerCost='LinearWeighted', prefetch=False, prefetchInFlight=2, prefetchLDSFactor=0, prefetchMixMemOps=False, loadScale_A='BufferToVGPR', loadScale_B='BufferToVGPR', swizzleScale=False, swizzleTileSize={'m': 0, 'k': 0, 'n': 0, 'l': 0}, prefetchScale=False, streamK=True, numWGs=256, streamKTwoTile=False, streamKTwoTileDPFirst=False, architecture={'ArchString': 'gfx950', 'Xnack': False, 'Sramecc': True}, matchMemoryAccess=True, version='')
+   0.24% | p=2.5347e-02 
	| 3. FloatsGEMM(M: 7680, N: 8448, K: 8448, alpha: 2, beta: 0.5, types: {'type_A': 'half', 'type_B': 'half', 'type_C': 'half', 'type_D': 'half', 'type_acc': 'float', 'trans_A': 'N', 'trans_B': 'T', 'scale_A': 'None', 'scaleType_A': 'None', 'scale_B': 'None', 'scaleType_B': 'None', 'scaleBlockSize': -1, 'scaleSkipPermlane': False, 'scaleShuffleTileA': [], 'scaleShuffleTileB': []}, scaleValue_A: 1, scaleValue_B: 1, initMode_A: DataInitMode(Bounded), initMode_B: DataInitMode(Bounded), initMode_C: DataInitMode(Bounded), workgroupMappingDim: -1, workgroupMappingValue: -1)| GEMM(mac_m: 128, mac_n: 128, mac_k: 16, wave_m: 32, wave_n: 32, wave_k: 8, wave_b: 1, workgroup_size_x: 128, workgroup_size_y: 2, workgroupRemapXCC: False, workgroupRemapXCCValue: -1, unroll_x: 0, unroll_y: 0, load_A: BufferToLDSViaVGPR, load_B: BufferToLDSViaVGPR, storeLDS_D: True, betaInFma: True, scheduler: Priority, schedulerCost: LinearWeighted, prefetch: False, prefetchInFlight: 0, prefetchLDSFactor: 0, prefetchMixMemOps: False, loadScale_A: BufferToVGPR, loadScale_B: BufferToVGPR, swizzleScale: False, swizzleTileSize: {'m': 0, 'k': 0, 'n': 0, 'l': 0}, prefetchScale: False, streamK: True, numWGs: 256, streamKTwoTile: False, streamKTwoTileDPFirst: False, architecture: {'ArchString': 'gfx950', 'Xnack': False, 'Sramecc': True}, matchMemoryAccess: True, version: ) | GEMMProblem(M=7680, N=8448, K=8448, alpha=2, beta=0.5, types={'type_A': 'half', 'type_B': 'half', 'type_C': 'half', 'type_D': 'half', 'type_acc': 'float', 'trans_A': 'N', 'trans_B': 'T', 'scale_A': 'None', 'scaleType_A': 'None', 'scale_B': 'None', 'scaleType_B': 'None', 'scaleBlockSize': -1, 'scaleSkipPermlane': False, 'scaleShuffleTileA': [], 'scaleShuffleTileB': []}, scaleValue_A=1, scaleValue_B=1, initMode_A='DataInitMode(Bounded)', initMode_B='DataInitMode(Bounded)', initMode_C='DataInitMode(Bounded)', workgroupMappingDim=-1, workgroupMappingValue=-1)GEMMSolution(mac_m=128, mac_n=128, mac_k=16, wave_m=32, wave_n=32, wave_k=8, wave_b=1, workgroup_size_x=128, workgroup_size_y=2, workgroupRemapXCC=False, workgroupRemapXCCValue=-1, unroll_x=0, unroll_y=0, load_A='BufferToLDSViaVGPR', load_B='BufferToLDSViaVGPR', storeLDS_D=True, betaInFma=True, scheduler='Priority', schedulerCost='LinearWeighted', prefetch=False, prefetchInFlight=0, prefetchLDSFactor=0, prefetchMixMemOps=False, loadScale_A='BufferToVGPR', loadScale_B='BufferToVGPR', swizzleScale=False, swizzleTileSize={'m': 0, 'k': 0, 'n': 0, 'l': 0}, prefetchScale=False, streamK=True, numWGs=256, streamKTwoTile=False, streamKTwoTileDPFirst=False, architecture={'ArchString': 'gfx950', 'Xnack': False, 'Sramecc': True}, matchMemoryAccess=True, version='')
-   0.51% | p=2.5347e-02 
	| 3. FloatsGEMM(M: 7680, N: 8448, K: 8448, alpha: 2, beta: 0.5, types: {'type_A': 'half', 'type_B': 'half', 'type_C': 'half', 'type_D': 'half', 'type_acc': 'float', 'trans_A': 'N', 'trans_B': 'T', 'scale_A': 'None', 'scaleType_A': 'None', 'scale_B': 'None', 'scaleType_B': 'None', 'scaleBlockSize': -1, 'scaleSkipPermlane': False, 'scaleShuffleTileA': [], 'scaleShuffleTileB': []}, scaleValue_A: 1, scaleValue_B: 1, initMode_A: DataInitMode(Bounded), initMode_B: DataInitMode(Bounded), initMode_C: DataInitMode(Bounded), workgroupMappingDim: -1, workgroupMappingValue: -1)| GEMM(mac_m: 128, mac_n: 128, mac_k: 32, wave_m: 32, wave_n: 32, wave_k: 8, wave_b: 1, workgroup_size_x: 128, workgroup_size_y: 2, workgroupRemapXCC: False, workgroupRemapXCCValue: -1, unroll_x: 0, unroll_y: 0, load_A: BufferToLDSViaVGPR, load_B: BufferToLDSViaVGPR, storeLDS_D: True, betaInFma: True, scheduler: Priority, schedulerCost: LinearWeighted, prefetch: False, prefetchInFlight: 0, prefetchLDSFactor: 0, prefetchMixMemOps: False, loadScale_A: BufferToVGPR, loadScale_B: BufferToVGPR, swizzleScale: False, swizzleTileSize: {'m': 0, 'k': 0, 'n': 0, 'l': 0}, prefetchScale: False, streamK: True, numWGs: 256, streamKTwoTile: False, streamKTwoTileDPFirst: True, architecture: {'ArchString': 'gfx950', 'Xnack': False, 'Sramecc': True}, matchMemoryAccess: True, version: ) | GEMMProblem(M=7680, N=8448, K=8448, alpha=2, beta=0.5, types={'type_A': 'half', 'type_B': 'half', 'type_C': 'half', 'type_D': 'half', 'type_acc': 'float', 'trans_A': 'N', 'trans_B': 'T', 'scale_A': 'None', 'scaleType_A': 'None', 'scale_B': 'None', 'scaleType_B': 'None', 'scaleBlockSize': -1, 'scaleSkipPermlane': False, 'scaleShuffleTileA': [], 'scaleShuffleTileB': []}, scaleValue_A=1, scaleValue_B=1, initMode_A='DataInitMode(Bounded)', initMode_B='DataInitMode(Bounded)', initMode_C='DataInitMode(Bounded)', workgroupMappingDim=-1, workgroupMappingValue=-1)GEMMSolution(mac_m=128, mac_n=128, mac_k=32, wave_m=32, wave_n=32, wave_k=8, wave_b=1, workgroup_size_x=128, workgroup_size_y=2, workgroupRemapXCC=False, workgroupRemapXCCValue=-1, unroll_x=0, unroll_y=0, load_A='BufferToLDSViaVGPR', load_B='BufferToLDSViaVGPR', storeLDS_D=True, betaInFma=True, scheduler='Priority', schedulerCost='LinearWeighted', prefetch=False, prefetchInFlight=0, prefetchLDSFactor=0, prefetchMixMemOps=False, loadScale_A='BufferToVGPR', loadScale_B='BufferToVGPR', swizzleScale=False, swizzleTileSize={'m': 0, 'k': 0, 'n': 0, 'l': 0}, prefetchScale=False, streamK=True, numWGs=256, streamKTwoTile=False, streamKTwoTileDPFirst=True, architecture={'ArchString': 'gfx950', 'Xnack': False, 'Sramecc': True}, matchMemoryAccess=True, version='')
+   7.19% | p=5.6994e-05 
	| 3. FloatsGEMM(M: 7680, N: 8448, K: 8448, alpha: 2, beta: 0.5, types: {'type_A': 'half', 'type_B': 'half', 'type_C': 'half', 'type_D': 'half', 'type_acc': 'float', 'trans_A': 'N', 'trans_B': 'T', 'scale_A': 'None', 'scaleType_A': 'None', 'scale_B': 'None', 'scaleType_B': 'None', 'scaleBlockSize': -1, 'scaleSkipPermlane': False, 'scaleShuffleTileA': [], 'scaleShuffleTileB': []}, scaleValue_A: 1, scaleValue_B: 1, initMode_A: DataInitMode(Bounded), initMode_B: DataInitMode(Bounded), initMode_C: DataInitMode(Bounded), workgroupMappingDim: -1, workgroupMappingValue: -1)| GEMM(mac_m: 128, mac_n: 64, mac_k: 16, wave_m: 32, wave_n: 32, wave_k: 8, wave_b: 1, workgroup_size_x: 128, workgroup_size_y: 2, workgroupRemapXCC: False, workgroupRemapXCCValue: -1, unroll_x: 0, unroll_y: 0, load_A: BufferToLDSViaVGPR, load_B: BufferToLDSViaVGPR, storeLDS_D: True, betaInFma: True, scheduler: Priority, schedulerCost: LinearWeighted, prefetch: False, prefetchInFlight: 0, prefetchLDSFactor: 0, prefetchMixMemOps: False, loadScale_A: BufferToVGPR, loadScale_B: BufferToVGPR, swizzleScale: False, swizzleTileSize: {'m': 0, 'k': 0, 'n': 0, 'l': 0}, prefetchScale: False, streamK: True, numWGs: 256, streamKTwoTile: True, streamKTwoTileDPFirst: False, architecture: {'ArchString': 'gfx950', 'Xnack': False, 'Sramecc': True}, matchMemoryAccess: True, version: ) | GEMMProblem(M=7680, N=8448, K=8448, alpha=2, beta=0.5, types={'type_A': 'half', 'type_B': 'half', 'type_C': 'half', 'type_D': 'half', 'type_acc': 'float', 'trans_A': 'N', 'trans_B': 'T', 'scale_A': 'None', 'scaleType_A': 'None', 'scale_B': 'None', 'scaleType_B': 'None', 'scaleBlockSize': -1, 'scaleSkipPermlane': False, 'scaleShuffleTileA': [], 'scaleShuffleTileB': []}, scaleValue_A=1, scaleValue_B=1, initMode_A='DataInitMode(Bounded)', initMode_B='DataInitMode(Bounded)', initMode_C='DataInitMode(Bounded)', workgroupMappingDim=-1, workgroupMappingValue=-1)GEMMSolution(mac_m=128, mac_n=64, mac_k=16, wave_m=32, wave_n=32, wave_k=8, wave_b=1, workgroup_size_x=128, workgroup_size_y=2, workgroupRemapXCC=False, workgroupRemapXCCValue=-1, unroll_x=0, unroll_y=0, load_A='BufferToLDSViaVGPR', load_B='BufferToLDSViaVGPR', storeLDS_D=True, betaInFma=True, scheduler='Priority', schedulerCost='LinearWeighted', prefetch=False, prefetchInFlight=0, prefetchLDSFactor=0, prefetchMixMemOps=False, loadScale_A='BufferToVGPR', loadScale_B='BufferToVGPR', swizzleScale=False, swizzleTileSize={'m': 0, 'k': 0, 'n': 0, 'l': 0}, prefetchScale=False, streamK=True, numWGs=256, streamKTwoTile=True, streamKTwoTileDPFirst=False, architecture={'ArchString': 'gfx950', 'Xnack': False, 'Sramecc': True}, matchMemoryAccess=True, version='')
+   0.40% | p=2.5347e-02 
	| 3. FloatsGEMM(M: 7680, N: 8448, K: 8448, alpha: 2, beta: 0.5, types: {'type_A': 'half', 'type_B': 'half', 'type_C': 'half', 'type_D': 'half', 'type_acc': 'float', 'trans_A': 'N', 'trans_B': 'T', 'scale_A': 'None', 'scaleType_A': 'None', 'scale_B': 'None', 'scaleType_B': 'None', 'scaleBlockSize': -1, 'scaleSkipPermlane': False, 'scaleShuffleTileA': [], 'scaleShuffleTileB': []}, scaleValue_A: 1, scaleValue_B: 1, initMode_A: DataInitMode(Bounded), initMode_B: DataInitMode(Bounded), initMode_C: DataInitMode(Bounded), workgroupMappingDim: -1, workgroupMappingValue: -1)| GEMM(mac_m: 128, mac_n: 64, mac_k: 32, wave_m: 32, wave_n: 32, wave_k: 8, wave_b: 1, workgroup_size_x: 128, workgroup_size_y: 2, workgroupRemapXCC: False, workgroupRemapXCCValue: -1, unroll_x: 0, unroll_y: 0, load_A: BufferToLDSViaVGPR, load_B: BufferToLDSViaVGPR, storeLDS_D: True, betaInFma: True, scheduler: Priority, schedulerCost: LinearWeighted, prefetch: False, prefetchInFlight: 0, prefetchLDSFactor: 0, prefetchMixMemOps: False, loadScale_A: BufferToVGPR, loadScale_B: BufferToVGPR, swizzleScale: False, swizzleTileSize: {'m': 0, 'k': 0, 'n': 0, 'l': 0}, prefetchScale: False, streamK: True, numWGs: 256, streamKTwoTile: False, streamKTwoTileDPFirst: False, architecture: {'ArchString': 'gfx950', 'Xnack': False, 'Sramecc': True}, matchMemoryAccess: True, version: ) | GEMMProblem(M=7680, N=8448, K=8448, alpha=2, beta=0.5, types={'type_A': 'half', 'type_B': 'half', 'type_C': 'half', 'type_D': 'half', 'type_acc': 'float', 'trans_A': 'N', 'trans_B': 'T', 'scale_A': 'None', 'scaleType_A': 'None', 'scale_B': 'None', 'scaleType_B': 'None', 'scaleBlockSize': -1, 'scaleSkipPermlane': False, 'scaleShuffleTileA': [], 'scaleShuffleTileB': []}, scaleValue_A=1, scaleValue_B=1, initMode_A='DataInitMode(Bounded)', initMode_B='DataInitMode(Bounded)', initMode_C='DataInitMode(Bounded)', workgroupMappingDim=-1, workgroupMappingValue=-1)GEMMSolution(mac_m=128, mac_n=64, mac_k=32, wave_m=32, wave_n=32, wave_k=8, wave_b=1, workgroup_size_x=128, workgroup_size_y=2, workgroupRemapXCC=False, workgroupRemapXCCValue=-1, unroll_x=0, unroll_y=0, load_A='BufferToLDSViaVGPR', load_B='BufferToLDSViaVGPR', storeLDS_D=True, betaInFma=True, scheduler='Priority', schedulerCost='LinearWeighted', prefetch=False, prefetchInFlight=0, prefetchLDSFactor=0, prefetchMixMemOps=False, loadScale_A='BufferToVGPR', loadScale_B='BufferToVGPR', swizzleScale=False, swizzleTileSize={'m': 0, 'k': 0, 'n': 0, 'l': 0}, prefetchScale=False, streamK=True, numWGs=256, streamKTwoTile=False, streamKTwoTileDPFirst=False, architecture={'ArchString': 'gfx950', 'Xnack': False, 'Sramecc': True}, matchMemoryAccess=True, version='')
+   0.95% | p=5.6994e-05 
	| 3. FloatsGEMM(M: 7680, N: 8448, K: 8448, alpha: 2, beta: 0.5, types: {'type_A': 'half', 'type_B': 'half', 'type_C': 'half', 'type_D': 'half', 'type_acc': 'float', 'trans_A': 'N', 'trans_B': 'T', 'scale_A': 'None', 'scaleType_A': 'None', 'scale_B': 'None', 'scaleType_B': 'None', 'scaleBlockSize': -1, 'scaleSkipPermlane': False, 'scaleShuffleTileA': [], 'scaleShuffleTileB': []}, scaleValue_A: 1, scaleValue_B: 1, initMode_A: DataInitMode(Bounded), initMode_B: DataInitMode(Bounded), initMode_C: DataInitMode(Bounded), workgroupMappingDim: -1, workgroupMappingValue: -1)| GEMM(mac_m: 128, mac_n: 64, mac_k: 32, wave_m: 32, wave_n: 32, wave_k: 8, wave_b: 1, workgroup_size_x: 128, workgroup_size_y: 2, workgroupRemapXCC: False, workgroupRemapXCCValue: -1, unroll_x: 0, unroll_y: 0, load_A: BufferToLDSViaVGPR, load_B: BufferToLDSViaVGPR, storeLDS_D: True, betaInFma: True, scheduler: Priority, schedulerCost: LinearWeighted, prefetch: False, prefetchInFlight: 0, prefetchLDSFactor: 0, prefetchMixMemOps: False, loadScale_A: BufferToVGPR, loadScale_B: BufferToVGPR, swizzleScale: False, swizzleTileSize: {'m': 0, 'k': 0, 'n': 0, 'l': 0}, prefetchScale: False, streamK: True, numWGs: 256, streamKTwoTile: True, streamKTwoTileDPFirst: False, architecture: {'ArchString': 'gfx950', 'Xnack': False, 'Sramecc': True}, matchMemoryAccess: True, version: ) | GEMMProblem(M=7680, N=8448, K=8448, alpha=2, beta=0.5, types={'type_A': 'half', 'type_B': 'half', 'type_C': 'half', 'type_D': 'half', 'type_acc': 'float', 'trans_A': 'N', 'trans_B': 'T', 'scale_A': 'None', 'scaleType_A': 'None', 'scale_B': 'None', 'scaleType_B': 'None', 'scaleBlockSize': -1, 'scaleSkipPermlane': False, 'scaleShuffleTileA': [], 'scaleShuffleTileB': []}, scaleValue_A=1, scaleValue_B=1, initMode_A='DataInitMode(Bounded)', initMode_B='DataInitMode(Bounded)', initMode_C='DataInitMode(Bounded)', workgroupMappingDim=-1, workgroupMappingValue=-1)GEMMSolution(mac_m=128, mac_n=64, mac_k=32, wave_m=32, wave_n=32, wave_k=8, wave_b=1, workgroup_size_x=128, workgroup_size_y=2, workgroupRemapXCC=False, workgroupRemapXCCValue=-1, unroll_x=0, unroll_y=0, load_A='BufferToLDSViaVGPR', load_B='BufferToLDSViaVGPR', storeLDS_D=True, betaInFma=True, scheduler='Priority', schedulerCost='LinearWeighted', prefetch=False, prefetchInFlight=0, prefetchLDSFactor=0, prefetchMixMemOps=False, loadScale_A='BufferToVGPR', loadScale_B='BufferToVGPR', swizzleScale=False, swizzleTileSize={'m': 0, 'k': 0, 'n': 0, 'l': 0}, prefetchScale=False, streamK=True, numWGs=256, streamKTwoTile=True, streamKTwoTileDPFirst=False, architecture={'ArchString': 'gfx950', 'Xnack': False, 'Sramecc': True}, matchMemoryAccess=True, version='')
+   0.28% | p=2.5347e-02 
	| 3. FloatsGEMM(M: 7680, N: 8448, K: 8448, alpha: 2, beta: 0.5, types: {'type_A': 'half', 'type_B': 'half', 'type_C': 'half', 'type_D': 'half', 'type_acc': 'float', 'trans_A': 'N', 'trans_B': 'T', 'scale_A': 'None', 'scaleType_A': 'None', 'scale_B': 'None', 'scaleType_B': 'None', 'scaleBlockSize': -1, 'scaleSkipPermlane': False, 'scaleShuffleTileA': [], 'scaleShuffleTileB': []}, scaleValue_A: 1, scaleValue_B: 1, initMode_A: DataInitMode(Bounded), initMode_B: DataInitMode(Bounded), initMode_C: DataInitMode(Bounded), workgroupMappingDim: -1, workgroupMappingValue: -1)| GEMM(mac_m: 128, mac_n: 64, mac_k: 64, wave_m: 32, wave_n: 32, wave_k: 8, wave_b: 1, workgroup_size_x: 128, workgroup_size_y: 2, workgroupRemapXCC: False, workgroupRemapXCCValue: -1, unroll_x: 0, unroll_y: 0, load_A: BufferToLDSViaVGPR, load_B: BufferToLDSViaVGPR, storeLDS_D: True, betaInFma: True, scheduler: Priority, schedulerCost: LinearWeighted, prefetch: False, prefetchInFlight: 0, prefetchLDSFactor: 0, prefetchMixMemOps: False, loadScale_A: BufferToVGPR, loadScale_B: BufferToVGPR, swizzleScale: False, swizzleTileSize: {'m': 0, 'k': 0, 'n': 0, 'l': 0}, prefetchScale: False, streamK: True, numWGs: 256, streamKTwoTile: True, streamKTwoTileDPFirst: False, architecture: {'ArchString': 'gfx950', 'Xnack': False, 'Sramecc': True}, matchMemoryAccess: True, version: ) | GEMMProblem(M=7680, N=8448, K=8448, alpha=2, beta=0.5, types={'type_A': 'half', 'type_B': 'half', 'type_C': 'half', 'type_D': 'half', 'type_acc': 'float', 'trans_A': 'N', 'trans_B': 'T', 'scale_A': 'None', 'scaleType_A': 'None', 'scale_B': 'None', 'scaleType_B': 'None', 'scaleBlockSize': -1, 'scaleSkipPermlane': False, 'scaleShuffleTileA': [], 'scaleShuffleTileB': []}, scaleValue_A=1, scaleValue_B=1, initMode_A='DataInitMode(Bounded)', initMode_B='DataInitMode(Bounded)', initMode_C='DataInitMode(Bounded)', workgroupMappingDim=-1, workgroupMappingValue=-1)GEMMSolution(mac_m=128, mac_n=64, mac_k=64, wave_m=32, wave_n=32, wave_k=8, wave_b=1, workgroup_size_x=128, workgroup_size_y=2, workgroupRemapXCC=False, workgroupRemapXCCValue=-1, unroll_x=0, unroll_y=0, load_A='BufferToLDSViaVGPR', load_B='BufferToLDSViaVGPR', storeLDS_D=True, betaInFma=True, scheduler='Priority', schedulerCost='LinearWeighted', prefetch=False, prefetchInFlight=0, prefetchLDSFactor=0, prefetchMixMemOps=False, loadScale_A='BufferToVGPR', loadScale_B='BufferToVGPR', swizzleScale=False, swizzleTileSize={'m': 0, 'k': 0, 'n': 0, 'l': 0}, prefetchScale=False, streamK=True, numWGs=256, streamKTwoTile=True, streamKTwoTileDPFirst=False, architecture={'ArchString': 'gfx950', 'Xnack': False, 'Sramecc': True}, matchMemoryAccess=True, version='')
-   0.93% | p=1.7451e-03 
	| 3. FloatsGEMM(M: 7680, N: 8448, K: 8448, alpha: 2, beta: 0.5, types: {'type_A': 'half', 'type_B': 'half', 'type_C': 'half', 'type_D': 'half', 'type_acc': 'float', 'trans_A': 'N', 'trans_B': 'T', 'scale_A': 'None', 'scaleType_A': 'None', 'scale_B': 'None', 'scaleType_B': 'None', 'scaleBlockSize': -1, 'scaleSkipPermlane': False, 'scaleShuffleTileA': [], 'scaleShuffleTileB': []}, scaleValue_A: 1, scaleValue_B: 1, initMode_A: DataInitMode(Bounded), initMode_B: DataInitMode(Bounded), initMode_C: DataInitMode(Bounded), workgroupMappingDim: -1, workgroupMappingValue: -1)| GEMM(mac_m: 64, mac_n: 128, mac_k: 16, wave_m: 32, wave_n: 32, wave_k: 8, wave_b: 1, workgroup_size_x: 128, workgroup_size_y: 2, workgroupRemapXCC: False, workgroupRemapXCCValue: -1, unroll_x: 0, unroll_y: 0, load_A: BufferToLDSViaVGPR, load_B: BufferToLDSViaVGPR, storeLDS_D: True, betaInFma: True, scheduler: Priority, schedulerCost: LinearWeighted, prefetch: False, prefetchInFlight: 0, prefetchLDSFactor: 0, prefetchMixMemOps: False, loadScale_A: BufferToVGPR, loadScale_B: BufferToVGPR, swizzleScale: False, swizzleTileSize: {'m': 0, 'k': 0, 'n': 0, 'l': 0}, prefetchScale: False, streamK: True, numWGs: 256, streamKTwoTile: False, streamKTwoTileDPFirst: False, architecture: {'ArchString': 'gfx950', 'Xnack': False, 'Sramecc': True}, matchMemoryAccess: True, version: ) | GEMMProblem(M=7680, N=8448, K=8448, alpha=2, beta=0.5, types={'type_A': 'half', 'type_B': 'half', 'type_C': 'half', 'type_D': 'half', 'type_acc': 'float', 'trans_A': 'N', 'trans_B': 'T', 'scale_A': 'None', 'scaleType_A': 'None', 'scale_B': 'None', 'scaleType_B': 'None', 'scaleBlockSize': -1, 'scaleSkipPermlane': False, 'scaleShuffleTileA': [], 'scaleShuffleTileB': []}, scaleValue_A=1, scaleValue_B=1, initMode_A='DataInitMode(Bounded)', initMode_B='DataInitMode(Bounded)', initMode_C='DataInitMode(Bounded)', workgroupMappingDim=-1, workgroupMappingValue=-1)GEMMSolution(mac_m=64, mac_n=128, mac_k=16, wave_m=32, wave_n=32, wave_k=8, wave_b=1, workgroup_size_x=128, workgroup_size_y=2, workgroupRemapXCC=False, workgroupRemapXCCValue=-1, unroll_x=0, unroll_y=0, load_A='BufferToLDSViaVGPR', load_B='BufferToLDSViaVGPR', storeLDS_D=True, betaInFma=True, scheduler='Priority', schedulerCost='LinearWeighted', prefetch=False, prefetchInFlight=0, prefetchLDSFactor=0, prefetchMixMemOps=False, loadScale_A='BufferToVGPR', loadScale_B='BufferToVGPR', swizzleScale=False, swizzleTileSize={'m': 0, 'k': 0, 'n': 0, 'l': 0}, prefetchScale=False, streamK=True, numWGs=256, streamKTwoTile=False, streamKTwoTileDPFirst=False, architecture={'ArchString': 'gfx950', 'Xnack': False, 'Sramecc': True}, matchMemoryAccess=True, version='')
+   0.30% | p=2.5347e-02 
	| 3. FloatsGEMM(M: 7680, N: 8448, K: 8448, alpha: 2, beta: 0.5, types: {'type_A': 'half', 'type_B': 'half', 'type_C': 'half', 'type_D': 'half', 'type_acc': 'float', 'trans_A': 'N', 'trans_B': 'T', 'scale_A': 'None', 'scaleType_A': 'None', 'scale_B': 'None', 'scaleType_B': 'None', 'scaleBlockSize': -1, 'scaleSkipPermlane': False, 'scaleShuffleTileA': [], 'scaleShuffleTileB': []}, scaleValue_A: 1, scaleValue_B: 1, initMode_A: DataInitMode(Bounded), initMode_B: DataInitMode(Bounded), initMode_C: DataInitMode(Bounded), workgroupMappingDim: -1, workgroupMappingValue: -1)| GEMM(mac_m: 64, mac_n: 128, mac_k: 16, wave_m: 32, wave_n: 32, wave_k: 8, wave_b: 1, workgroup_size_x: 128, workgroup_size_y: 2, workgroupRemapXCC: False, workgroupRemapXCCValue: -1, unroll_x: 0, unroll_y: 0, load_A: BufferToLDSViaVGPR, load_B: BufferToLDSViaVGPR, storeLDS_D: True, betaInFma: True, scheduler: Priority, schedulerCost: LinearWeighted, prefetch: False, prefetchInFlight: 0, prefetchLDSFactor: 0, prefetchMixMemOps: False, loadScale_A: BufferToVGPR, loadScale_B: BufferToVGPR, swizzleScale: False, swizzleTileSize: {'m': 0, 'k': 0, 'n': 0, 'l': 0}, prefetchScale: False, streamK: True, numWGs: 256, streamKTwoTile: False, streamKTwoTileDPFirst: True, architecture: {'ArchString': 'gfx950', 'Xnack': False, 'Sramecc': True}, matchMemoryAccess: True, version: ) | GEMMProblem(M=7680, N=8448, K=8448, alpha=2, beta=0.5, types={'type_A': 'half', 'type_B': 'half', 'type_C': 'half', 'type_D': 'half', 'type_acc': 'float', 'trans_A': 'N', 'trans_B': 'T', 'scale_A': 'None', 'scaleType_A': 'None', 'scale_B': 'None', 'scaleType_B': 'None', 'scaleBlockSize': -1, 'scaleSkipPermlane': False, 'scaleShuffleTileA': [], 'scaleShuffleTileB': []}, scaleValue_A=1, scaleValue_B=1, initMode_A='DataInitMode(Bounded)', initMode_B='DataInitMode(Bounded)', initMode_C='DataInitMode(Bounded)', workgroupMappingDim=-1, workgroupMappingValue=-1)GEMMSolution(mac_m=64, mac_n=128, mac_k=16, wave_m=32, wave_n=32, wave_k=8, wave_b=1, workgroup_size_x=128, workgroup_size_y=2, workgroupRemapXCC=False, workgroupRemapXCCValue=-1, unroll_x=0, unroll_y=0, load_A='BufferToLDSViaVGPR', load_B='BufferToLDSViaVGPR', storeLDS_D=True, betaInFma=True, scheduler='Priority', schedulerCost='LinearWeighted', prefetch=False, prefetchInFlight=0, prefetchLDSFactor=0, prefetchMixMemOps=False, loadScale_A='BufferToVGPR', loadScale_B='BufferToVGPR', swizzleScale=False, swizzleTileSize={'m': 0, 'k': 0, 'n': 0, 'l': 0}, prefetchScale=False, streamK=True, numWGs=256, streamKTwoTile=False, streamKTwoTileDPFirst=True, architecture={'ArchString': 'gfx950', 'Xnack': False, 'Sramecc': True}, matchMemoryAccess=True, version='')
+   0.77% | p=1.7451e-03 
	| 3. FloatsGEMM(M: 7680, N: 8448, K: 8448, alpha: 2, beta: 0.5, types: {'type_A': 'half', 'type_B': 'half', 'type_C': 'half', 'type_D': 'half', 'type_acc': 'float', 'trans_A': 'N', 'trans_B': 'T', 'scale_A': 'None', 'scaleType_A': 'None', 'scale_B': 'None', 'scaleType_B': 'None', 'scaleBlockSize': -1, 'scaleSkipPermlane': False, 'scaleShuffleTileA': [], 'scaleShuffleTileB': []}, scaleValue_A: 1, scaleValue_B: 1, initMode_A: DataInitMode(Bounded), initMode_B: DataInitMode(Bounded), initMode_C: DataInitMode(Bounded), workgroupMappingDim: -1, workgroupMappingValue: -1)| GEMM(mac_m: 64, mac_n: 128, mac_k: 16, wave_m: 32, wave_n: 32, wave_k: 8, wave_b: 1, workgroup_size_x: 128, workgroup_size_y: 2, workgroupRemapXCC: False, workgroupRemapXCCValue: -1, unroll_x: 0, unroll_y: 0, load_A: BufferToLDSViaVGPR, load_B: BufferToLDSViaVGPR, storeLDS_D: True, betaInFma: True, scheduler: Priority, schedulerCost: LinearWeighted, prefetch: True, prefetchInFlight: 2, prefetchLDSFactor: 2, prefetchMixMemOps: False, loadScale_A: BufferToVGPR, loadScale_B: BufferToVGPR, swizzleScale: False, swizzleTileSize: {'m': 0, 'k': 0, 'n': 0, 'l': 0}, prefetchScale: False, streamK: True, numWGs: 256, streamKTwoTile: False, streamKTwoTileDPFirst: False, architecture: {'ArchString': 'gfx950', 'Xnack': False, 'Sramecc': True}, matchMemoryAccess: True, version: ) | GEMMProblem(M=7680, N=8448, K=8448, alpha=2, beta=0.5, types={'type_A': 'half', 'type_B': 'half', 'type_C': 'half', 'type_D': 'half', 'type_acc': 'float', 'trans_A': 'N', 'trans_B': 'T', 'scale_A': 'None', 'scaleType_A': 'None', 'scale_B': 'None', 'scaleType_B': 'None', 'scaleBlockSize': -1, 'scaleSkipPermlane': False, 'scaleShuffleTileA': [], 'scaleShuffleTileB': []}, scaleValue_A=1, scaleValue_B=1, initMode_A='DataInitMode(Bounded)', initMode_B='DataInitMode(Bounded)', initMode_C='DataInitMode(Bounded)', workgroupMappingDim=-1, workgroupMappingValue=-1)GEMMSolution(mac_m=64, mac_n=128, mac_k=16, wave_m=32, wave_n=32, wave_k=8, wave_b=1, workgroup_size_x=128, workgroup_size_y=2, workgroupRemapXCC=False, workgroupRemapXCCValue=-1, unroll_x=0, unroll_y=0, load_A='BufferToLDSViaVGPR', load_B='BufferToLDSViaVGPR', storeLDS_D=True, betaInFma=True, scheduler='Priority', schedulerCost='LinearWeighted', prefetch=True, prefetchInFlight=2, prefetchLDSFactor=2, prefetchMixMemOps=False, loadScale_A='BufferToVGPR', loadScale_B='BufferToVGPR', swizzleScale=False, swizzleTileSize={'m': 0, 'k': 0, 'n': 0, 'l': 0}, prefetchScale=False, streamK=True, numWGs=256, streamKTwoTile=False, streamKTwoTileDPFirst=False, architecture={'ArchString': 'gfx950', 'Xnack': False, 'Sramecc': True}, matchMemoryAccess=True, version='')
-   0.78% | p=1.7451e-03 
	| 3. FloatsGEMM(M: 7680, N: 8448, K: 8448, alpha: 2, beta: 0.5, types: {'type_A': 'half', 'type_B': 'half', 'type_C': 'half', 'type_D': 'half', 'type_acc': 'float', 'trans_A': 'N', 'trans_B': 'T', 'scale_A': 'None', 'scaleType_A': 'None', 'scale_B': 'None', 'scaleType_B': 'None', 'scaleBlockSize': -1, 'scaleSkipPermlane': False, 'scaleShuffleTileA': [], 'scaleShuffleTileB': []}, scaleValue_A: 1, scaleValue_B: 1, initMode_A: DataInitMode(Bounded), initMode_B: DataInitMode(Bounded), initMode_C: DataInitMode(Bounded), workgroupMappingDim: -1, workgroupMappingValue: -1)| GEMM(mac_m: 64, mac_n: 128, mac_k: 32, wave_m: 32, wave_n: 32, wave_k: 8, wave_b: 1, workgroup_size_x: 128, workgroup_size_y: 2, workgroupRemapXCC: False, workgroupRemapXCCValue: -1, unroll_x: 0, unroll_y: 0, load_A: BufferToLDSViaVGPR, load_B: BufferToLDSViaVGPR, storeLDS_D: True, betaInFma: True, scheduler: Priority, schedulerCost: LinearWeighted, prefetch: False, prefetchInFlight: 0, prefetchLDSFactor: 0, prefetchMixMemOps: False, loadScale_A: BufferToVGPR, loadScale_B: BufferToVGPR, swizzleScale: False, swizzleTileSize: {'m': 0, 'k': 0, 'n': 0, 'l': 0}, prefetchScale: False, streamK: True, numWGs: 256, streamKTwoTile: False, streamKTwoTileDPFirst: False, architecture: {'ArchString': 'gfx950', 'Xnack': False, 'Sramecc': True}, matchMemoryAccess: True, version: ) | GEMMProblem(M=7680, N=8448, K=8448, alpha=2, beta=0.5, types={'type_A': 'half', 'type_B': 'half', 'type_C': 'half', 'type_D': 'half', 'type_acc': 'float', 'trans_A': 'N', 'trans_B': 'T', 'scale_A': 'None', 'scaleType_A': 'None', 'scale_B': 'None', 'scaleType_B': 'None', 'scaleBlockSize': -1, 'scaleSkipPermlane': False, 'scaleShuffleTileA': [], 'scaleShuffleTileB': []}, scaleValue_A=1, scaleValue_B=1, initMode_A='DataInitMode(Bounded)', initMode_B='DataInitMode(Bounded)', initMode_C='DataInitMode(Bounded)', workgroupMappingDim=-1, workgroupMappingValue=-1)GEMMSolution(mac_m=64, mac_n=128, mac_k=32, wave_m=32, wave_n=32, wave_k=8, wave_b=1, workgroup_size_x=128, workgroup_size_y=2, workgroupRemapXCC=False, workgroupRemapXCCValue=-1, unroll_x=0, unroll_y=0, load_A='BufferToLDSViaVGPR', load_B='BufferToLDSViaVGPR', storeLDS_D=True, betaInFma=True, scheduler='Priority', schedulerCost='LinearWeighted', prefetch=False, prefetchInFlight=0, prefetchLDSFactor=0, prefetchMixMemOps=False, loadScale_A='BufferToVGPR', loadScale_B='BufferToVGPR', swizzleScale=False, swizzleTileSize={'m': 0, 'k': 0, 'n': 0, 'l': 0}, prefetchScale=False, streamK=True, numWGs=256, streamKTwoTile=False, streamKTwoTileDPFirst=False, architecture={'ArchString': 'gfx950', 'Xnack': False, 'Sramecc': True}, matchMemoryAccess=True, version='')
+   0.59% | p=1.7451e-03 
	| 3. FloatsGEMM(M: 7680, N: 8448, K: 8448, alpha: 2, beta: 0.5, types: {'type_A': 'half', 'type_B': 'half', 'type_C': 'half', 'type_D': 'half', 'type_acc': 'float', 'trans_A': 'N', 'trans_B': 'T', 'scale_A': 'None', 'scaleType_A': 'None', 'scale_B': 'None', 'scaleType_B': 'None', 'scaleBlockSize': -1, 'scaleSkipPermlane': False, 'scaleShuffleTileA': [], 'scaleShuffleTileB': []}, scaleValue_A: 1, scaleValue_B: 1, initMode_A: DataInitMode(Bounded), initMode_B: DataInitMode(Bounded), initMode_C: DataInitMode(Bounded), workgroupMappingDim: -1, workgroupMappingValue: -1)| GEMM(mac_m: 64, mac_n: 128, mac_k: 32, wave_m: 32, wave_n: 32, wave_k: 8, wave_b: 1, workgroup_size_x: 128, workgroup_size_y: 2, workgroupRemapXCC: False, workgroupRemapXCCValue: -1, unroll_x: 0, unroll_y: 0, load_A: BufferToLDSViaVGPR, load_B: BufferToLDSViaVGPR, storeLDS_D: True, betaInFma: True, scheduler: Priority, schedulerCost: LinearWeighted, prefetch: False, prefetchInFlight: 0, prefetchLDSFactor: 0, prefetchMixMemOps: False, loadScale_A: BufferToVGPR, loadScale_B: BufferToVGPR, swizzleScale: False, swizzleTileSize: {'m': 0, 'k': 0, 'n': 0, 'l': 0}, prefetchScale: False, streamK: True, numWGs: 256, streamKTwoTile: True, streamKTwoTileDPFirst: False, architecture: {'ArchString': 'gfx950', 'Xnack': False, 'Sramecc': True}, matchMemoryAccess: True, version: ) | GEMMProblem(M=7680, N=8448, K=8448, alpha=2, beta=0.5, types={'type_A': 'half', 'type_B': 'half', 'type_C': 'half', 'type_D': 'half', 'type_acc': 'float', 'trans_A': 'N', 'trans_B': 'T', 'scale_A': 'None', 'scaleType_A': 'None', 'scale_B': 'None', 'scaleType_B': 'None', 'scaleBlockSize': -1, 'scaleSkipPermlane': False, 'scaleShuffleTileA': [], 'scaleShuffleTileB': []}, scaleValue_A=1, scaleValue_B=1, initMode_A='DataInitMode(Bounded)', initMode_B='DataInitMode(Bounded)', initMode_C='DataInitMode(Bounded)', workgroupMappingDim=-1, workgroupMappingValue=-1)GEMMSolution(mac_m=64, mac_n=128, mac_k=32, wave_m=32, wave_n=32, wave_k=8, wave_b=1, workgroup_size_x=128, workgroup_size_y=2, workgroupRemapXCC=False, workgroupRemapXCCValue=-1, unroll_x=0, unroll_y=0, load_A='BufferToLDSViaVGPR', load_B='BufferToLDSViaVGPR', storeLDS_D=True, betaInFma=True, scheduler='Priority', schedulerCost='LinearWeighted', prefetch=False, prefetchInFlight=0, prefetchLDSFactor=0, prefetchMixMemOps=False, loadScale_A='BufferToVGPR', loadScale_B='BufferToVGPR', swizzleScale=False, swizzleTileSize={'m': 0, 'k': 0, 'n': 0, 'l': 0}, prefetchScale=False, streamK=True, numWGs=256, streamKTwoTile=True, streamKTwoTileDPFirst=False, architecture={'ArchString': 'gfx950', 'Xnack': False, 'Sramecc': True}, matchMemoryAccess=True, version='')
+   0.24% | p=1.7451e-03 
	| 3. FloatsGEMM(M: 7680, N: 8448, K: 8448, alpha: 2, beta: 0.5, types: {'type_A': 'half', 'type_B': 'half', 'type_C': 'half', 'type_D': 'half', 'type_acc': 'float', 'trans_A': 'N', 'trans_B': 'T', 'scale_A': 'None', 'scaleType_A': 'None', 'scale_B': 'None', 'scaleType_B': 'None', 'scaleBlockSize': -1, 'scaleSkipPermlane': False, 'scaleShuffleTileA': [], 'scaleShuffleTileB': []}, scaleValue_A: 1, scaleValue_B: 1, initMode_A: DataInitMode(Bounded), initMode_B: DataInitMode(Bounded), initMode_C: DataInitMode(Bounded), workgroupMappingDim: -1, workgroupMappingValue: -1)| GEMM(mac_m: 64, mac_n: 128, mac_k: 64, wave_m: 32, wave_n: 32, wave_k: 8, wave_b: 1, workgroup_size_x: 128, workgroup_size_y: 2, workgroupRemapXCC: False, workgroupRemapXCCValue: -1, unroll_x: 0, unroll_y: 0, load_A: BufferToLDSViaVGPR, load_B: BufferToLDSViaVGPR, storeLDS_D: True, betaInFma: True, scheduler: Priority, schedulerCost: LinearWeighted, prefetch: False, prefetchInFlight: 0, prefetchLDSFactor: 0, prefetchMixMemOps: False, loadScale_A: BufferToVGPR, loadScale_B: BufferToVGPR, swizzleScale: False, swizzleTileSize: {'m': 0, 'k': 0, 'n': 0, 'l': 0}, prefetchScale: False, streamK: True, numWGs: 256, streamKTwoTile: False, streamKTwoTileDPFirst: False, architecture: {'ArchString': 'gfx950', 'Xnack': False, 'Sramecc': True}, matchMemoryAccess: True, version: ) | GEMMProblem(M=7680, N=8448, K=8448, alpha=2, beta=0.5, types={'type_A': 'half', 'type_B': 'half', 'type_C': 'half', 'type_D': 'half', 'type_acc': 'float', 'trans_A': 'N', 'trans_B': 'T', 'scale_A': 'None', 'scaleType_A': 'None', 'scale_B': 'None', 'scaleType_B': 'None', 'scaleBlockSize': -1, 'scaleSkipPermlane': False, 'scaleShuffleTileA': [], 'scaleShuffleTileB': []}, scaleValue_A=1, scaleValue_B=1, initMode_A='DataInitMode(Bounded)', initMode_B='DataInitMode(Bounded)', initMode_C='DataInitMode(Bounded)', workgroupMappingDim=-1, workgroupMappingValue=-1)GEMMSolution(mac_m=64, mac_n=128, mac_k=64, wave_m=32, wave_n=32, wave_k=8, wave_b=1, workgroup_size_x=128, workgroup_size_y=2, workgroupRemapXCC=False, workgroupRemapXCCValue=-1, unroll_x=0, unroll_y=0, load_A='BufferToLDSViaVGPR', load_B='BufferToLDSViaVGPR', storeLDS_D=True, betaInFma=True, scheduler='Priority', schedulerCost='LinearWeighted', prefetch=False, prefetchInFlight=0, prefetchLDSFactor=0, prefetchMixMemOps=False, loadScale_A='BufferToVGPR', loadScale_B='BufferToVGPR', swizzleScale=False, swizzleTileSize={'m': 0, 'k': 0, 'n': 0, 'l': 0}, prefetchScale=False, streamK=True, numWGs=256, streamKTwoTile=False, streamKTwoTileDPFirst=False, architecture={'ArchString': 'gfx950', 'Xnack': False, 'Sramecc': True}, matchMemoryAccess=True, version='')
-   0.60% | p=2.5347e-02 
	| 3. FloatsGEMM(M: 7680, N: 8448, K: 8448, alpha: 2, beta: 0.5, types: {'type_A': 'half', 'type_B': 'half', 'type_C': 'half', 'type_D': 'half', 'type_acc': 'float', 'trans_A': 'N', 'trans_B': 'T', 'scale_A': 'None', 'scaleType_A': 'None', 'scale_B': 'None', 'scaleType_B': 'None', 'scaleBlockSize': -1, 'scaleSkipPermlane': False, 'scaleShuffleTileA': [], 'scaleShuffleTileB': []}, scaleValue_A: 1, scaleValue_B: 1, initMode_A: DataInitMode(Bounded), initMode_B: DataInitMode(Bounded), initMode_C: DataInitMode(Bounded), workgroupMappingDim: -1, workgroupMappingValue: -1)| GEMM(mac_m: 64, mac_n: 256, mac_k: 16, wave_m: 32, wave_n: 32, wave_k: 8, wave_b: 1, workgroup_size_x: 128, workgroup_size_y: 2, workgroupRemapXCC: False, workgroupRemapXCCValue: -1, unroll_x: 0, unroll_y: 0, load_A: BufferToLDSViaVGPR, load_B: BufferToLDSViaVGPR, storeLDS_D: True, betaInFma: True, scheduler: Priority, schedulerCost: LinearWeighted, prefetch: False, prefetchInFlight: 0, prefetchLDSFactor: 0, prefetchMixMemOps: False, loadScale_A: BufferToVGPR, loadScale_B: BufferToVGPR, swizzleScale: False, swizzleTileSize: {'m': 0, 'k': 0, 'n': 0, 'l': 0}, prefetchScale: False, streamK: True, numWGs: 256, streamKTwoTile: False, streamKTwoTileDPFirst: False, architecture: {'ArchString': 'gfx950', 'Xnack': False, 'Sramecc': True}, matchMemoryAccess: True, version: ) | GEMMProblem(M=7680, N=8448, K=8448, alpha=2, beta=0.5, types={'type_A': 'half', 'type_B': 'half', 'type_C': 'half', 'type_D': 'half', 'type_acc': 'float', 'trans_A': 'N', 'trans_B': 'T', 'scale_A': 'None', 'scaleType_A': 'None', 'scale_B': 'None', 'scaleType_B': 'None', 'scaleBlockSize': -1, 'scaleSkipPermlane': False, 'scaleShuffleTileA': [], 'scaleShuffleTileB': []}, scaleValue_A=1, scaleValue_B=1, initMode_A='DataInitMode(Bounded)', initMode_B='DataInitMode(Bounded)', initMode_C='DataInitMode(Bounded)', workgroupMappingDim=-1, workgroupMappingValue=-1)GEMMSolution(mac_m=64, mac_n=256, mac_k=16, wave_m=32, wave_n=32, wave_k=8, wave_b=1, workgroup_size_x=128, workgroup_size_y=2, workgroupRemapXCC=False, workgroupRemapXCCValue=-1, unroll_x=0, unroll_y=0, load_A='BufferToLDSViaVGPR', load_B='BufferToLDSViaVGPR', storeLDS_D=True, betaInFma=True, scheduler='Priority', schedulerCost='LinearWeighted', prefetch=False, prefetchInFlight=0, prefetchLDSFactor=0, prefetchMixMemOps=False, loadScale_A='BufferToVGPR', loadScale_B='BufferToVGPR', swizzleScale=False, swizzleTileSize={'m': 0, 'k': 0, 'n': 0, 'l': 0}, prefetchScale=False, streamK=True, numWGs=256, streamKTwoTile=False, streamKTwoTileDPFirst=False, architecture={'ArchString': 'gfx950', 'Xnack': False, 'Sramecc': True}, matchMemoryAccess=True, version='')
-   0.56% | p=1.7451e-03 
	| 3. FloatsGEMM(M: 7680, N: 8448, K: 8448, alpha: 2, beta: 0.5, types: {'type_A': 'half', 'type_B': 'half', 'type_C': 'half', 'type_D': 'half', 'type_acc': 'float', 'trans_A': 'N', 'trans_B': 'T', 'scale_A': 'None', 'scaleType_A': 'None', 'scale_B': 'None', 'scaleType_B': 'None', 'scaleBlockSize': -1, 'scaleSkipPermlane': False, 'scaleShuffleTileA': [], 'scaleShuffleTileB': []}, scaleValue_A: 1, scaleValue_B: 1, initMode_A: DataInitMode(Bounded), initMode_B: DataInitMode(Bounded), initMode_C: DataInitMode(Bounded), workgroupMappingDim: -1, workgroupMappingValue: -1)| GEMM(mac_m: 64, mac_n: 256, mac_k: 16, wave_m: 32, wave_n: 32, wave_k: 8, wave_b: 1, workgroup_size_x: 128, workgroup_size_y: 2, workgroupRemapXCC: False, workgroupRemapXCCValue: -1, unroll_x: 0, unroll_y: 0, load_A: BufferToLDSViaVGPR, load_B: BufferToLDSViaVGPR, storeLDS_D: True, betaInFma: True, scheduler: Priority, schedulerCost: LinearWeighted, prefetch: False, prefetchInFlight: 0, prefetchLDSFactor: 0, prefetchMixMemOps: False, loadScale_A: BufferToVGPR, loadScale_B: BufferToVGPR, swizzleScale: False, swizzleTileSize: {'m': 0, 'k': 0, 'n': 0, 'l': 0}, prefetchScale: False, streamK: True, numWGs: 256, streamKTwoTile: False, streamKTwoTileDPFirst: True, architecture: {'ArchString': 'gfx950', 'Xnack': False, 'Sramecc': True}, matchMemoryAccess: True, version: ) | GEMMProblem(M=7680, N=8448, K=8448, alpha=2, beta=0.5, types={'type_A': 'half', 'type_B': 'half', 'type_C': 'half', 'type_D': 'half', 'type_acc': 'float', 'trans_A': 'N', 'trans_B': 'T', 'scale_A': 'None', 'scaleType_A': 'None', 'scale_B': 'None', 'scaleType_B': 'None', 'scaleBlockSize': -1, 'scaleSkipPermlane': False, 'scaleShuffleTileA': [], 'scaleShuffleTileB': []}, scaleValue_A=1, scaleValue_B=1, initMode_A='DataInitMode(Bounded)', initMode_B='DataInitMode(Bounded)', initMode_C='DataInitMode(Bounded)', workgroupMappingDim=-1, workgroupMappingValue=-1)GEMMSolution(mac_m=64, mac_n=256, mac_k=16, wave_m=32, wave_n=32, wave_k=8, wave_b=1, workgroup_size_x=128, workgroup_size_y=2, workgroupRemapXCC=False, workgroupRemapXCCValue=-1, unroll_x=0, unroll_y=0, load_A='BufferToLDSViaVGPR', load_B='BufferToLDSViaVGPR', storeLDS_D=True, betaInFma=True, scheduler='Priority', schedulerCost='LinearWeighted', prefetch=False, prefetchInFlight=0, prefetchLDSFactor=0, prefetchMixMemOps=False, loadScale_A='BufferToVGPR', loadScale_B='BufferToVGPR', swizzleScale=False, swizzleTileSize={'m': 0, 'k': 0, 'n': 0, 'l': 0}, prefetchScale=False, streamK=True, numWGs=256, streamKTwoTile=False, streamKTwoTileDPFirst=True, architecture={'ArchString': 'gfx950', 'Xnack': False, 'Sramecc': True}, matchMemoryAccess=True, version='')
-   1.16% | p=5.6994e-05 
	| 3. FloatsGEMM(M: 7680, N: 8448, K: 8448, alpha: 2, beta: 0.5, types: {'type_A': 'half', 'type_B': 'half', 'type_C': 'half', 'type_D': 'half', 'type_acc': 'float', 'trans_A': 'N', 'trans_B': 'T', 'scale_A': 'None', 'scaleType_A': 'None', 'scale_B': 'None', 'scaleType_B': 'None', 'scaleBlockSize': -1, 'scaleSkipPermlane': False, 'scaleShuffleTileA': [], 'scaleShuffleTileB': []}, scaleValue_A: 1, scaleValue_B: 1, initMode_A: DataInitMode(Bounded), initMode_B: DataInitMode(Bounded), initMode_C: DataInitMode(Bounded), workgroupMappingDim: -1, workgroupMappingValue: -1)| GEMM(mac_m: 64, mac_n: 256, mac_k: 16, wave_m: 32, wave_n: 32, wave_k: 8, wave_b: 1, workgroup_size_x: 128, workgroup_size_y: 2, workgroupRemapXCC: False, workgroupRemapXCCValue: -1, unroll_x: 0, unroll_y: 0, load_A: BufferToLDSViaVGPR, load_B: BufferToLDSViaVGPR, storeLDS_D: True, betaInFma: True, scheduler: Priority, schedulerCost: LinearWeighted, prefetch: False, prefetchInFlight: 0, prefetchLDSFactor: 0, prefetchMixMemOps: False, loadScale_A: BufferToVGPR, loadScale_B: BufferToVGPR, swizzleScale: False, swizzleTileSize: {'m': 0, 'k': 0, 'n': 0, 'l': 0}, prefetchScale: False, streamK: True, numWGs: 256, streamKTwoTile: True, streamKTwoTileDPFirst: False, architecture: {'ArchString': 'gfx950', 'Xnack': False, 'Sramecc': True}, matchMemoryAccess: True, version: ) | GEMMProblem(M=7680, N=8448, K=8448, alpha=2, beta=0.5, types={'type_A': 'half', 'type_B': 'half', 'type_C': 'half', 'type_D': 'half', 'type_acc': 'float', 'trans_A': 'N', 'trans_B': 'T', 'scale_A': 'None', 'scaleType_A': 'None', 'scale_B': 'None', 'scaleType_B': 'None', 'scaleBlockSize': -1, 'scaleSkipPermlane': False, 'scaleShuffleTileA': [], 'scaleShuffleTileB': []}, scaleValue_A=1, scaleValue_B=1, initMode_A='DataInitMode(Bounded)', initMode_B='DataInitMode(Bounded)', initMode_C='DataInitMode(Bounded)', workgroupMappingDim=-1, workgroupMappingValue=-1)GEMMSolution(mac_m=64, mac_n=256, mac_k=16, wave_m=32, wave_n=32, wave_k=8, wave_b=1, workgroup_size_x=128, workgroup_size_y=2, workgroupRemapXCC=False, workgroupRemapXCCValue=-1, unroll_x=0, unroll_y=0, load_A='BufferToLDSViaVGPR', load_B='BufferToLDSViaVGPR', storeLDS_D=True, betaInFma=True, scheduler='Priority', schedulerCost='LinearWeighted', prefetch=False, prefetchInFlight=0, prefetchLDSFactor=0, prefetchMixMemOps=False, loadScale_A='BufferToVGPR', loadScale_B='BufferToVGPR', swizzleScale=False, swizzleTileSize={'m': 0, 'k': 0, 'n': 0, 'l': 0}, prefetchScale=False, streamK=True, numWGs=256, streamKTwoTile=True, streamKTwoTileDPFirst=False, architecture={'ArchString': 'gfx950', 'Xnack': False, 'Sramecc': True}, matchMemoryAccess=True, version='')
+   2.22% | p=2.5347e-02 
	| 3. FloatsGEMM(M: 7680, N: 8448, K: 8448, alpha: 2, beta: 0.5, types: {'type_A': 'half', 'type_B': 'half', 'type_C': 'half', 'type_D': 'half', 'type_acc': 'float', 'trans_A': 'N', 'trans_B': 'T', 'scale_A': 'None', 'scaleType_A': 'None', 'scale_B': 'None', 'scaleType_B': 'None', 'scaleBlockSize': -1, 'scaleSkipPermlane': False, 'scaleShuffleTileA': [], 'scaleShuffleTileB': []}, scaleValue_A: 1, scaleValue_B: 1, initMode_A: DataInitMode(Bounded), initMode_B: DataInitMode(Bounded), initMode_C: DataInitMode(Bounded), workgroupMappingDim: -1, workgroupMappingValue: -1)| GEMM(mac_m: 64, mac_n: 64, mac_k: 16, wave_m: 32, wave_n: 32, wave_k: 8, wave_b: 1, workgroup_size_x: 128, workgroup_size_y: 2, workgroupRemapXCC: False, workgroupRemapXCCValue: -1, unroll_x: 0, unroll_y: 0, load_A: BufferToLDSViaVGPR, load_B: BufferToLDSViaVGPR, storeLDS_D: True, betaInFma: True, scheduler: Priority, schedulerCost: LinearWeighted, prefetch: False, prefetchInFlight: 0, prefetchLDSFactor: 0, prefetchMixMemOps: False, loadScale_A: BufferToVGPR, loadScale_B: BufferToVGPR, swizzleScale: False, swizzleTileSize: {'m': 0, 'k': 0, 'n': 0, 'l': 0}, prefetchScale: False, streamK: True, numWGs: 256, streamKTwoTile: True, streamKTwoTileDPFirst: False, architecture: {'ArchString': 'gfx950', 'Xnack': False, 'Sramecc': True}, matchMemoryAccess: True, version: ) | GEMMProblem(M=7680, N=8448, K=8448, alpha=2, beta=0.5, types={'type_A': 'half', 'type_B': 'half', 'type_C': 'half', 'type_D': 'half', 'type_acc': 'float', 'trans_A': 'N', 'trans_B': 'T', 'scale_A': 'None', 'scaleType_A': 'None', 'scale_B': 'None', 'scaleType_B': 'None', 'scaleBlockSize': -1, 'scaleSkipPermlane': False, 'scaleShuffleTileA': [], 'scaleShuffleTileB': []}, scaleValue_A=1, scaleValue_B=1, initMode_A='DataInitMode(Bounded)', initMode_B='DataInitMode(Bounded)', initMode_C='DataInitMode(Bounded)', workgroupMappingDim=-1, workgroupMappingValue=-1)GEMMSolution(mac_m=64, mac_n=64, mac_k=16, wave_m=32, wave_n=32, wave_k=8, wave_b=1, workgroup_size_x=128, workgroup_size_y=2, workgroupRemapXCC=False, workgroupRemapXCCValue=-1, unroll_x=0, unroll_y=0, load_A='BufferToLDSViaVGPR', load_B='BufferToLDSViaVGPR', storeLDS_D=True, betaInFma=True, scheduler='Priority', schedulerCost='LinearWeighted', prefetch=False, prefetchInFlight=0, prefetchLDSFactor=0, prefetchMixMemOps=False, loadScale_A='BufferToVGPR', loadScale_B='BufferToVGPR', swizzleScale=False, swizzleTileSize={'m': 0, 'k': 0, 'n': 0, 'l': 0}, prefetchScale=False, streamK=True, numWGs=256, streamKTwoTile=True, streamKTwoTileDPFirst=False, architecture={'ArchString': 'gfx950', 'Xnack': False, 'Sramecc': True}, matchMemoryAccess=True, version='')
-   0.48% | p=1.7451e-03 
	| 3. FloatsGEMM(M: 7680, N: 8448, K: 8448, alpha: 2, beta: 0.5, types: {'type_A': 'half', 'type_B': 'half', 'type_C': 'half', 'type_D': 'half', 'type_acc': 'float', 'trans_A': 'N', 'trans_B': 'T', 'scale_A': 'None', 'scaleType_A': 'None', 'scale_B': 'None', 'scaleType_B': 'None', 'scaleBlockSize': -1, 'scaleSkipPermlane': False, 'scaleShuffleTileA': [], 'scaleShuffleTileB': []}, scaleValue_A: 1, scaleValue_B: 1, initMode_A: DataInitMode(Bounded), initMode_B: DataInitMode(Bounded), initMode_C: DataInitMode(Bounded), workgroupMappingDim: -1, workgroupMappingValue: -1)| GEMM(mac_m: 64, mac_n: 64, mac_k: 16, wave_m: 32, wave_n: 32, wave_k: 8, wave_b: 1, workgroup_size_x: 128, workgroup_size_y: 2, workgroupRemapXCC: False, workgroupRemapXCCValue: -1, unroll_x: 0, unroll_y: 0, load_A: BufferToLDSViaVGPR, load_B: BufferToLDSViaVGPR, storeLDS_D: True, betaInFma: True, scheduler: Priority, schedulerCost: LinearWeighted, prefetch: True, prefetchInFlight: 2, prefetchLDSFactor: 2, prefetchMixMemOps: False, loadScale_A: BufferToVGPR, loadScale_B: BufferToVGPR, swizzleScale: False, swizzleTileSize: {'m': 0, 'k': 0, 'n': 0, 'l': 0}, prefetchScale: False, streamK: True, numWGs: 256, streamKTwoTile: False, streamKTwoTileDPFirst: False, architecture: {'ArchString': 'gfx950', 'Xnack': False, 'Sramecc': True}, matchMemoryAccess: True, version: ) | GEMMProblem(M=7680, N=8448, K=8448, alpha=2, beta=0.5, types={'type_A': 'half', 'type_B': 'half', 'type_C': 'half', 'type_D': 'half', 'type_acc': 'float', 'trans_A': 'N', 'trans_B': 'T', 'scale_A': 'None', 'scaleType_A': 'None', 'scale_B': 'None', 'scaleType_B': 'None', 'scaleBlockSize': -1, 'scaleSkipPermlane': False, 'scaleShuffleTileA': [], 'scaleShuffleTileB': []}, scaleValue_A=1, scaleValue_B=1, initMode_A='DataInitMode(Bounded)', initMode_B='DataInitMode(Bounded)', initMode_C='DataInitMode(Bounded)', workgroupMappingDim=-1, workgroupMappingValue=-1)GEMMSolution(mac_m=64, mac_n=64, mac_k=16, wave_m=32, wave_n=32, wave_k=8, wave_b=1, workgroup_size_x=128, workgroup_size_y=2, workgroupRemapXCC=False, workgroupRemapXCCValue=-1, unroll_x=0, unroll_y=0, load_A='BufferToLDSViaVGPR', load_B='BufferToLDSViaVGPR', storeLDS_D=True, betaInFma=True, scheduler='Priority', schedulerCost='LinearWeighted', prefetch=True, prefetchInFlight=2, prefetchLDSFactor=2, prefetchMixMemOps=False, loadScale_A='BufferToVGPR', loadScale_B='BufferToVGPR', swizzleScale=False, swizzleTileSize={'m': 0, 'k': 0, 'n': 0, 'l': 0}, prefetchScale=False, streamK=True, numWGs=256, streamKTwoTile=False, streamKTwoTileDPFirst=False, architecture={'ArchString': 'gfx950', 'Xnack': False, 'Sramecc': True}, matchMemoryAccess=True, version='')
-   0.70% | p=2.5347e-02 
	| 3. FloatsGEMM(M: 7680, N: 8448, K: 8448, alpha: 2, beta: 0.5, types: {'type_A': 'half', 'type_B': 'half', 'type_C': 'half', 'type_D': 'half', 'type_acc': 'float', 'trans_A': 'N', 'trans_B': 'T', 'scale_A': 'None', 'scaleType_A': 'None', 'scale_B': 'None', 'scaleType_B': 'None', 'scaleBlockSize': -1, 'scaleSkipPermlane': False, 'scaleShuffleTileA': [], 'scaleShuffleTileB': []}, scaleValue_A: 1, scaleValue_B: 1, initMode_A: DataInitMode(Bounded), initMode_B: DataInitMode(Bounded), initMode_C: DataInitMode(Bounded), workgroupMappingDim: -1, workgroupMappingValue: -1)| GEMM(mac_m: 64, mac_n: 64, mac_k: 32, wave_m: 32, wave_n: 32, wave_k: 8, wave_b: 1, workgroup_size_x: 128, workgroup_size_y: 2, workgroupRemapXCC: False, workgroupRemapXCCValue: -1, unroll_x: 0, unroll_y: 0, load_A: BufferToLDSViaVGPR, load_B: BufferToLDSViaVGPR, storeLDS_D: True, betaInFma: True, scheduler: Priority, schedulerCost: LinearWeighted, prefetch: False, prefetchInFlight: 0, prefetchLDSFactor: 0, prefetchMixMemOps: False, loadScale_A: BufferToVGPR, loadScale_B: BufferToVGPR, swizzleScale: False, swizzleTileSize: {'m': 0, 'k': 0, 'n': 0, 'l': 0}, prefetchScale: False, streamK: True, numWGs: 256, streamKTwoTile: False, streamKTwoTileDPFirst: False, architecture: {'ArchString': 'gfx950', 'Xnack': False, 'Sramecc': True}, matchMemoryAccess: True, version: ) | GEMMProblem(M=7680, N=8448, K=8448, alpha=2, beta=0.5, types={'type_A': 'half', 'type_B': 'half', 'type_C': 'half', 'type_D': 'half', 'type_acc': 'float', 'trans_A': 'N', 'trans_B': 'T', 'scale_A': 'None', 'scaleType_A': 'None', 'scale_B': 'None', 'scaleType_B': 'None', 'scaleBlockSize': -1, 'scaleSkipPermlane': False, 'scaleShuffleTileA': [], 'scaleShuffleTileB': []}, scaleValue_A=1, scaleValue_B=1, initMode_A='DataInitMode(Bounded)', initMode_B='DataInitMode(Bounded)', initMode_C='DataInitMode(Bounded)', workgroupMappingDim=-1, workgroupMappingValue=-1)GEMMSolution(mac_m=64, mac_n=64, mac_k=32, wave_m=32, wave_n=32, wave_k=8, wave_b=1, workgroup_size_x=128, workgroup_size_y=2, workgroupRemapXCC=False, workgroupRemapXCCValue=-1, unroll_x=0, unroll_y=0, load_A='BufferToLDSViaVGPR', load_B='BufferToLDSViaVGPR', storeLDS_D=True, betaInFma=True, scheduler='Priority', schedulerCost='LinearWeighted', prefetch=False, prefetchInFlight=0, prefetchLDSFactor=0, prefetchMixMemOps=False, loadScale_A='BufferToVGPR', loadScale_B='BufferToVGPR', swizzleScale=False, swizzleTileSize={'m': 0, 'k': 0, 'n': 0, 'l': 0}, prefetchScale=False, streamK=True, numWGs=256, streamKTwoTile=False, streamKTwoTileDPFirst=False, architecture={'ArchString': 'gfx950', 'Xnack': False, 'Sramecc': True}, matchMemoryAccess=True, version='')
+   0.18% | p=2.5347e-02 
	| 3. FloatsGEMM(M: 7680, N: 8448, K: 8448, alpha: 2, beta: 0.5, types: {'type_A': 'half', 'type_B': 'half', 'type_C': 'half', 'type_D': 'half', 'type_acc': 'float', 'trans_A': 'N', 'trans_B': 'T', 'scale_A': 'None', 'scaleType_A': 'None', 'scale_B': 'None', 'scaleType_B': 'None', 'scaleBlockSize': -1, 'scaleSkipPermlane': False, 'scaleShuffleTileA': [], 'scaleShuffleTileB': []}, scaleValue_A: 1, scaleValue_B: 1, initMode_A: DataInitMode(Bounded), initMode_B: DataInitMode(Bounded), initMode_C: DataInitMode(Bounded), workgroupMappingDim: -1, workgroupMappingValue: -1)| GEMM(mac_m: 64, mac_n: 64, mac_k: 32, wave_m: 32, wave_n: 32, wave_k: 8, wave_b: 1, workgroup_size_x: 128, workgroup_size_y: 2, workgroupRemapXCC: False, workgroupRemapXCCValue: -1, unroll_x: 0, unroll_y: 0, load_A: BufferToLDSViaVGPR, load_B: BufferToLDSViaVGPR, storeLDS_D: True, betaInFma: True, scheduler: Priority, schedulerCost: LinearWeighted, prefetch: False, prefetchInFlight: 0, prefetchLDSFactor: 0, prefetchMixMemOps: False, loadScale_A: BufferToVGPR, loadScale_B: BufferToVGPR, swizzleScale: False, swizzleTileSize: {'m': 0, 'k': 0, 'n': 0, 'l': 0}, prefetchScale: False, streamK: True, numWGs: 256, streamKTwoTile: True, streamKTwoTileDPFirst: False, architecture: {'ArchString': 'gfx950', 'Xnack': False, 'Sramecc': True}, matchMemoryAccess: True, version: ) | GEMMProblem(M=7680, N=8448, K=8448, alpha=2, beta=0.5, types={'type_A': 'half', 'type_B': 'half', 'type_C': 'half', 'type_D': 'half', 'type_acc': 'float', 'trans_A': 'N', 'trans_B': 'T', 'scale_A': 'None', 'scaleType_A': 'None', 'scale_B': 'None', 'scaleType_B': 'None', 'scaleBlockSize': -1, 'scaleSkipPermlane': False, 'scaleShuffleTileA': [], 'scaleShuffleTileB': []}, scaleValue_A=1, scaleValue_B=1, initMode_A='DataInitMode(Bounded)', initMode_B='DataInitMode(Bounded)', initMode_C='DataInitMode(Bounded)', workgroupMappingDim=-1, workgroupMappingValue=-1)GEMMSolution(mac_m=64, mac_n=64, mac_k=32, wave_m=32, wave_n=32, wave_k=8, wave_b=1, workgroup_size_x=128, workgroup_size_y=2, workgroupRemapXCC=False, workgroupRemapXCCValue=-1, unroll_x=0, unroll_y=0, load_A='BufferToLDSViaVGPR', load_B='BufferToLDSViaVGPR', storeLDS_D=True, betaInFma=True, scheduler='Priority', schedulerCost='LinearWeighted', prefetch=False, prefetchInFlight=0, prefetchLDSFactor=0, prefetchMixMemOps=False, loadScale_A='BufferToVGPR', loadScale_B='BufferToVGPR', swizzleScale=False, swizzleTileSize={'m': 0, 'k': 0, 'n': 0, 'l': 0}, prefetchScale=False, streamK=True, numWGs=256, streamKTwoTile=True, streamKTwoTileDPFirst=False, architecture={'ArchString': 'gfx950', 'Xnack': False, 'Sramecc': True}, matchMemoryAccess=True, version='')
-   0.96% | p=5.6994e-05 
	| 3. FloatsGEMM(M: 7680, N: 8448, K: 8448, alpha: 2, beta: 0.5, types: {'type_A': 'half', 'type_B': 'half', 'type_C': 'half', 'type_D': 'half', 'type_acc': 'float', 'trans_A': 'N', 'trans_B': 'T', 'scale_A': 'None', 'scaleType_A': 'None', 'scale_B': 'None', 'scaleType_B': 'None', 'scaleBlockSize': -1, 'scaleSkipPermlane': False, 'scaleShuffleTileA': [], 'scaleShuffleTileB': []}, scaleValue_A: 1, scaleValue_B: 1, initMode_A: DataInitMode(Bounded), initMode_B: DataInitMode(Bounded), initMode_C: DataInitMode(Bounded), workgroupMappingDim: -1, workgroupMappingValue: -1)| GEMM(mac_m: 64, mac_n: 64, mac_k: 64, wave_m: 32, wave_n: 32, wave_k: 8, wave_b: 1, workgroup_size_x: 128, workgroup_size_y: 2, workgroupRemapXCC: False, workgroupRemapXCCValue: -1, unroll_x: 0, unroll_y: 0, load_A: BufferToLDSViaVGPR, load_B: BufferToLDSViaVGPR, storeLDS_D: True, betaInFma: True, scheduler: Priority, schedulerCost: LinearWeighted, prefetch: False, prefetchInFlight: 0, prefetchLDSFactor: 0, prefetchMixMemOps: False, loadScale_A: BufferToVGPR, loadScale_B: BufferToVGPR, swizzleScale: False, swizzleTileSize: {'m': 0, 'k': 0, 'n': 0, 'l': 0}, prefetchScale: False, streamK: True, numWGs: 256, streamKTwoTile: False, streamKTwoTileDPFirst: False, architecture: {'ArchString': 'gfx950', 'Xnack': False, 'Sramecc': True}, matchMemoryAccess: True, version: ) | GEMMProblem(M=7680, N=8448, K=8448, alpha=2, beta=0.5, types={'type_A': 'half', 'type_B': 'half', 'type_C': 'half', 'type_D': 'half', 'type_acc': 'float', 'trans_A': 'N', 'trans_B': 'T', 'scale_A': 'None', 'scaleType_A': 'None', 'scale_B': 'None', 'scaleType_B': 'None', 'scaleBlockSize': -1, 'scaleSkipPermlane': False, 'scaleShuffleTileA': [], 'scaleShuffleTileB': []}, scaleValue_A=1, scaleValue_B=1, initMode_A='DataInitMode(Bounded)', initMode_B='DataInitMode(Bounded)', initMode_C='DataInitMode(Bounded)', workgroupMappingDim=-1, workgroupMappingValue=-1)GEMMSolution(mac_m=64, mac_n=64, mac_k=64, wave_m=32, wave_n=32, wave_k=8, wave_b=1, workgroup_size_x=128, workgroup_size_y=2, workgroupRemapXCC=False, workgroupRemapXCCValue=-1, unroll_x=0, unroll_y=0, load_A='BufferToLDSViaVGPR', load_B='BufferToLDSViaVGPR', storeLDS_D=True, betaInFma=True, scheduler='Priority', schedulerCost='LinearWeighted', prefetch=False, prefetchInFlight=0, prefetchLDSFactor=0, prefetchMixMemOps=False, loadScale_A='BufferToVGPR', loadScale_B='BufferToVGPR', swizzleScale=False, swizzleTileSize={'m': 0, 'k': 0, 'n': 0, 'l': 0}, prefetchScale=False, streamK=True, numWGs=256, streamKTwoTile=False, streamKTwoTileDPFirst=False, architecture={'ArchString': 'gfx950', 'Xnack': False, 'Sramecc': True}, matchMemoryAccess=True, version='')
+   1.63% | p=1.7451e-03 
	| 3. FloatsGEMM(M: 7680, N: 8448, K: 8448, alpha: 2, beta: 0.5, types: {'type_A': 'half', 'type_B': 'half', 'type_C': 'half', 'type_D': 'half', 'type_acc': 'float', 'trans_A': 'N', 'trans_B': 'T', 'scale_A': 'None', 'scaleType_A': 'None', 'scale_B': 'None', 'scaleType_B': 'None', 'scaleBlockSize': -1, 'scaleSkipPermlane': False, 'scaleShuffleTileA': [], 'scaleShuffleTileB': []}, scaleValue_A: 1, scaleValue_B: 1, initMode_A: DataInitMode(Bounded), initMode_B: DataInitMode(Bounded), initMode_C: DataInitMode(Bounded), workgroupMappingDim: -1, workgroupMappingValue: -1)| GEMM(mac_m: 64, mac_n: 64, mac_k: 64, wave_m: 32, wave_n: 32, wave_k: 8, wave_b: 1, workgroup_size_x: 128, workgroup_size_y: 2, workgroupRemapXCC: False, workgroupRemapXCCValue: -1, unroll_x: 0, unroll_y: 0, load_A: BufferToLDSViaVGPR, load_B: BufferToLDSViaVGPR, storeLDS_D: True, betaInFma: True, scheduler: Priority, schedulerCost: LinearWeighted, prefetch: True, prefetchInFlight: 2, prefetchLDSFactor: 2, prefetchMixMemOps: False, loadScale_A: BufferToVGPR, loadScale_B: BufferToVGPR, swizzleScale: False, swizzleTileSize: {'m': 0, 'k': 0, 'n': 0, 'l': 0}, prefetchScale: False, streamK: True, numWGs: 256, streamKTwoTile: False, streamKTwoTileDPFirst: False, architecture: {'ArchString': 'gfx950', 'Xnack': False, 'Sramecc': True}, matchMemoryAccess: True, version: ) | GEMMProblem(M=7680, N=8448, K=8448, alpha=2, beta=0.5, types={'type_A': 'half', 'type_B': 'half', 'type_C': 'half', 'type_D': 'half', 'type_acc': 'float', 'trans_A': 'N', 'trans_B': 'T', 'scale_A': 'None', 'scaleType_A': 'None', 'scale_B': 'None', 'scaleType_B': 'None', 'scaleBlockSize': -1, 'scaleSkipPermlane': False, 'scaleShuffleTileA': [], 'scaleShuffleTileB': []}, scaleValue_A=1, scaleValue_B=1, initMode_A='DataInitMode(Bounded)', initMode_B='DataInitMode(Bounded)', initMode_C='DataInitMode(Bounded)', workgroupMappingDim=-1, workgroupMappingValue=-1)GEMMSolution(mac_m=64, mac_n=64, mac_k=64, wave_m=32, wave_n=32, wave_k=8, wave_b=1, workgroup_size_x=128, workgroup_size_y=2, workgroupRemapXCC=False, workgroupRemapXCCValue=-1, unroll_x=0, unroll_y=0, load_A='BufferToLDSViaVGPR', load_B='BufferToLDSViaVGPR', storeLDS_D=True, betaInFma=True, scheduler='Priority', schedulerCost='LinearWeighted', prefetch=True, prefetchInFlight=2, prefetchLDSFactor=2, prefetchMixMemOps=False, loadScale_A='BufferToVGPR', loadScale_B='BufferToVGPR', swizzleScale=False, swizzleTileSize={'m': 0, 'k': 0, 'n': 0, 'l': 0}, prefetchScale=False, streamK=True, numWGs=256, streamKTwoTile=False, streamKTwoTileDPFirst=False, architecture={'ArchString': 'gfx950', 'Xnack': False, 'Sramecc': True}, matchMemoryAccess=True, version='')
+   1.21% | p=1.7451e-03 
	| 3. FloatsGEMM(M: 8448, N: 8448, K: 128, alpha: 2, beta: 0.5, types: {'type_A': 'half', 'type_B': 'half', 'type_C': 'half', 'type_D': 'half', 'type_acc': 'float', 'trans_A': 'N', 'trans_B': 'T', 'scale_A': 'None', 'scaleType_A': 'None', 'scale_B': 'None', 'scaleType_B': 'None', 'scaleBlockSize': -1, 'scaleSkipPermlane': False, 'scaleShuffleTileA': [], 'scaleShuffleTileB': []}, scaleValue_A: 1, scaleValue_B: 1, initMode_A: DataInitMode(Bounded), initMode_B: DataInitMode(Bounded), initMode_C: DataInitMode(Bounded), workgroupMappingDim: -1, workgroupMappingValue: -1)| GEMM(mac_m: 128, mac_n: 256, mac_k: 16, wave_m: 32, wave_n: 32, wave_k: 8, wave_b: 1, workgroup_size_x: 128, workgroup_size_y: 2, workgroupRemapXCC: False, workgroupRemapXCCValue: -1, unroll_x: 0, unroll_y: 0, load_A: BufferToLDSViaVGPR, load_B: BufferToLDSViaVGPR, storeLDS_D: True, betaInFma: True, scheduler: Sequential, schedulerCost: LinearWeighted, prefetch: True, prefetchInFlight: 2, prefetchLDSFactor: 0, prefetchMixMemOps: False, loadScale_A: BufferToVGPR, loadScale_B: BufferToVGPR, swizzleScale: False, swizzleTileSize: {'m': 0, 'k': 0, 'n': 0, 'l': 0}, prefetchScale: False, streamK: False, numWGs: 0, streamKTwoTile: False, streamKTwoTileDPFirst: False, architecture: {'ArchString': 'gfx950', 'Xnack': False, 'Sramecc': True}, matchMemoryAccess: True, version: ) | GEMMProblem(M=8448, N=8448, K=128, alpha=2, beta=0.5, types={'type_A': 'half', 'type_B': 'half', 'type_C': 'half', 'type_D': 'half', 'type_acc': 'float', 'trans_A': 'N', 'trans_B': 'T', 'scale_A': 'None', 'scaleType_A': 'None', 'scale_B': 'None', 'scaleType_B': 'None', 'scaleBlockSize': -1, 'scaleSkipPermlane': False, 'scaleShuffleTileA': [], 'scaleShuffleTileB': []}, scaleValue_A=1, scaleValue_B=1, initMode_A='DataInitMode(Bounded)', initMode_B='DataInitMode(Bounded)', initMode_C='DataInitMode(Bounded)', workgroupMappingDim=-1, workgroupMappingValue=-1)GEMMSolution(mac_m=128, mac_n=256, mac_k=16, wave_m=32, wave_n=32, wave_k=8, wave_b=1, workgroup_size_x=128, workgroup_size_y=2, workgroupRemapXCC=False, workgroupRemapXCCValue=-1, unroll_x=0, unroll_y=0, load_A='BufferToLDSViaVGPR', load_B='BufferToLDSViaVGPR', storeLDS_D=True, betaInFma=True, scheduler='Sequential', schedulerCost='LinearWeighted', prefetch=True, prefetchInFlight=2, prefetchLDSFactor=0, prefetchMixMemOps=False, loadScale_A='BufferToVGPR', loadScale_B='BufferToVGPR', swizzleScale=False, swizzleTileSize={'m': 0, 'k': 0, 'n': 0, 'l': 0}, prefetchScale=False, streamK=False, numWGs=0, streamKTwoTile=False, streamKTwoTileDPFirst=False, architecture={'ArchString': 'gfx950', 'Xnack': False, 'Sramecc': True}, matchMemoryAccess=True, version='')
-   0.75% | p=5.6994e-05 
	| CodeGen(instCount: 40000, instructions: comments)| CodeGen() | CodeGen(instCount: 40000, instructions: comments)
-   1.37% | p=2.5347e-02 
	| CodeGen(instCount: 40000, instructions: complex_mi_with_coop)| CodeGen() | CodeGen(instCount: 40000, instructions: complex_mi_with_coop)

Links

math-ci · 2025-12-10T18:33:05Z

Resource Report for gfx950-rocroller

Results

Details

@@                    Resource Usage Changes                         @@
====================================================================================================
- SGPR: 60 -> 63 (+3)
	| 3. FloatsGEMM(M: 3072, N: 4096, K: 4096, alpha: 2, beta: 0.5, types: {'type_A': 'float', 'type_B': 'float', 'type_C': 'float', 'type_D': 'float', 'type_acc': 'float', 'trans_A': 'N', 'trans_B': 'T', 'scale_A': 'None', 'scaleType_A': 'None', 'scale_B': 'None', 'scaleType_B': 'None', 'scaleBlockSize': -1, 'scaleSkipPermlane': False, 'scaleShuffleTileA': [], 'scaleShuffleTileB': []}, scaleValue_A: 1, scaleValue_B: 1, initMode_A: DataInitMode(Bounded), initMode_B: DataInitMode(Bounded), initMode_C: DataInitMode(Bounded), workgroupMappingDim: -1, workgroupMappingValue: -1)| GEMM(mac_m: 64, mac_n: 64, mac_k: 64, wave_m: 32, wave_n: 32, wave_k: 2, wave_b: 1, workgroup_size_x: 128, workgroup_size_y: 2, workgroupRemapXCC: False, workgroupRemapXCCValue: -1, unroll_x: 0, unroll_y: 0, load_A: BufferToLDSViaVGPR, load_B: BufferToLDSViaVGPR, storeLDS_D: True, betaInFma: True, scheduler: Priority, schedulerCost: LinearWeighted, prefetch: False, prefetchInFlight: 2, prefetchLDSFactor: 0, prefetchMixMemOps: False, loadScale_A: BufferToVGPR, loadScale_B: BufferToVGPR, swizzleScale: False, swizzleTileSize: {'m': 0, 'k': 0, 'n': 0, 'l': 0}, prefetchScale: False, streamK: True, numWGs: 256, streamKTwoTile: False, streamKTwoTileDPFirst: False, architecture: {'ArchString': 'gfx950', 'Xnack': False, 'Sramecc': True}, matchMemoryAccess: True, version: ) | GEMMProblem(M=3072, N=4096, K=4096, alpha=2, beta=0.5, types={'type_A': 'float', 'type_B': 'float', 'type_C': 'float', 'type_D': 'float', 'type_acc': 'float', 'trans_A': 'N', 'trans_B': 'T', 'scale_A': 'None', 'scaleType_A': 'None', 'scale_B': 'None', 'scaleType_B': 'None', 'scaleBlockSize': -1, 'scaleSkipPermlane': False, 'scaleShuffleTileA': [], 'scaleShuffleTileB': []}, scaleValue_A=1, scaleValue_B=1, initMode_A='DataInitMode(Bounded)', initMode_B='DataInitMode(Bounded)', initMode_C='DataInitMode(Bounded)', workgroupMappingDim=-1, workgroupMappingValue=-1)GEMMSolution(mac_m=64, mac_n=64, mac_k=64, wave_m=32, wave_n=32, wave_k=2, wave_b=1, workgroup_size_x=128, workgroup_size_y=2, workgroupRemapXCC=False, workgroupRemapXCCValue=-1, unroll_x=0, unroll_y=0, load_A='BufferToLDSViaVGPR', load_B='BufferToLDSViaVGPR', storeLDS_D=True, betaInFma=True, scheduler='Priority', schedulerCost='LinearWeighted', prefetch=False, prefetchInFlight=2, prefetchLDSFactor=0, prefetchMixMemOps=False, loadScale_A='BufferToVGPR', loadScale_B='BufferToVGPR', swizzleScale=False, swizzleTileSize={'m': 0, 'k': 0, 'n': 0, 'l': 0}, prefetchScale=False, streamK=True, numWGs=256, streamKTwoTile=False, streamKTwoTileDPFirst=False, architecture={'ArchString': 'gfx950', 'Xnack': False, 'Sramecc': True}, matchMemoryAccess=True, version='')
- SGPR: 63 -> 64 (+1)
	| 3. FloatsGEMM(M: 3072, N: 4096, K: 4096, alpha: 2, beta: 0.5, types: {'type_A': 'float', 'type_B': 'float', 'type_C': 'float', 'type_D': 'float', 'type_acc': 'float', 'trans_A': 'N', 'trans_B': 'T', 'scale_A': 'None', 'scaleType_A': 'None', 'scale_B': 'None', 'scaleType_B': 'None', 'scaleBlockSize': -1, 'scaleSkipPermlane': False, 'scaleShuffleTileA': [], 'scaleShuffleTileB': []}, scaleValue_A: 1, scaleValue_B: 1, initMode_A: DataInitMode(Bounded), initMode_B: DataInitMode(Bounded), initMode_C: DataInitMode(Bounded), workgroupMappingDim: -1, workgroupMappingValue: -1)| GEMM(mac_m: 64, mac_n: 64, mac_k: 64, wave_m: 32, wave_n: 32, wave_k: 2, wave_b: 1, workgroup_size_x: 128, workgroup_size_y: 2, workgroupRemapXCC: False, workgroupRemapXCCValue: -1, unroll_x: 0, unroll_y: 0, load_A: BufferToLDSViaVGPR, load_B: BufferToLDSViaVGPR, storeLDS_D: True, betaInFma: True, scheduler: Priority, schedulerCost: LinearWeighted, prefetch: False, prefetchInFlight: 2, prefetchLDSFactor: 0, prefetchMixMemOps: False, loadScale_A: BufferToVGPR, loadScale_B: BufferToVGPR, swizzleScale: False, swizzleTileSize: {'m': 0, 'k': 0, 'n': 0, 'l': 0}, prefetchScale: False, streamK: True, numWGs: 256, streamKTwoTile: False, streamKTwoTileDPFirst: True, architecture: {'ArchString': 'gfx950', 'Xnack': False, 'Sramecc': True}, matchMemoryAccess: True, version: ) | GEMMProblem(M=3072, N=4096, K=4096, alpha=2, beta=0.5, types={'type_A': 'float', 'type_B': 'float', 'type_C': 'float', 'type_D': 'float', 'type_acc': 'float', 'trans_A': 'N', 'trans_B': 'T', 'scale_A': 'None', 'scaleType_A': 'None', 'scale_B': 'None', 'scaleType_B': 'None', 'scaleBlockSize': -1, 'scaleSkipPermlane': False, 'scaleShuffleTileA': [], 'scaleShuffleTileB': []}, scaleValue_A=1, scaleValue_B=1, initMode_A='DataInitMode(Bounded)', initMode_B='DataInitMode(Bounded)', initMode_C='DataInitMode(Bounded)', workgroupMappingDim=-1, workgroupMappingValue=-1)GEMMSolution(mac_m=64, mac_n=64, mac_k=64, wave_m=32, wave_n=32, wave_k=2, wave_b=1, workgroup_size_x=128, workgroup_size_y=2, workgroupRemapXCC=False, workgroupRemapXCCValue=-1, unroll_x=0, unroll_y=0, load_A='BufferToLDSViaVGPR', load_B='BufferToLDSViaVGPR', storeLDS_D=True, betaInFma=True, scheduler='Priority', schedulerCost='LinearWeighted', prefetch=False, prefetchInFlight=2, prefetchLDSFactor=0, prefetchMixMemOps=False, loadScale_A='BufferToVGPR', loadScale_B='BufferToVGPR', swizzleScale=False, swizzleTileSize={'m': 0, 'k': 0, 'n': 0, 'l': 0}, prefetchScale=False, streamK=True, numWGs=256, streamKTwoTile=False, streamKTwoTileDPFirst=True, architecture={'ArchString': 'gfx950', 'Xnack': False, 'Sramecc': True}, matchMemoryAccess=True, version='')
- SGPR: 63 -> 64 (+1)
	| 3. FloatsGEMM(M: 3072, N: 4096, K: 4096, alpha: 2, beta: 0.5, types: {'type_A': 'float', 'type_B': 'float', 'type_C': 'float', 'type_D': 'float', 'type_acc': 'float', 'trans_A': 'N', 'trans_B': 'T', 'scale_A': 'None', 'scaleType_A': 'None', 'scale_B': 'None', 'scaleType_B': 'None', 'scaleBlockSize': -1, 'scaleSkipPermlane': False, 'scaleShuffleTileA': [], 'scaleShuffleTileB': []}, scaleValue_A: 1, scaleValue_B: 1, initMode_A: DataInitMode(Bounded), initMode_B: DataInitMode(Bounded), initMode_C: DataInitMode(Bounded), workgroupMappingDim: -1, workgroupMappingValue: -1)| GEMM(mac_m: 64, mac_n: 64, mac_k: 64, wave_m: 32, wave_n: 32, wave_k: 2, wave_b: 1, workgroup_size_x: 128, workgroup_size_y: 2, workgroupRemapXCC: False, workgroupRemapXCCValue: -1, unroll_x: 0, unroll_y: 0, load_A: BufferToLDSViaVGPR, load_B: BufferToLDSViaVGPR, storeLDS_D: True, betaInFma: True, scheduler: Priority, schedulerCost: LinearWeighted, prefetch: False, prefetchInFlight: 2, prefetchLDSFactor: 0, prefetchMixMemOps: False, loadScale_A: BufferToVGPR, loadScale_B: BufferToVGPR, swizzleScale: False, swizzleTileSize: {'m': 0, 'k': 0, 'n': 0, 'l': 0}, prefetchScale: False, streamK: True, numWGs: 256, streamKTwoTile: True, streamKTwoTileDPFirst: False, architecture: {'ArchString': 'gfx950', 'Xnack': False, 'Sramecc': True}, matchMemoryAccess: True, version: ) | GEMMProblem(M=3072, N=4096, K=4096, alpha=2, beta=0.5, types={'type_A': 'float', 'type_B': 'float', 'type_C': 'float', 'type_D': 'float', 'type_acc': 'float', 'trans_A': 'N', 'trans_B': 'T', 'scale_A': 'None', 'scaleType_A': 'None', 'scale_B': 'None', 'scaleType_B': 'None', 'scaleBlockSize': -1, 'scaleSkipPermlane': False, 'scaleShuffleTileA': [], 'scaleShuffleTileB': []}, scaleValue_A=1, scaleValue_B=1, initMode_A='DataInitMode(Bounded)', initMode_B='DataInitMode(Bounded)', initMode_C='DataInitMode(Bounded)', workgroupMappingDim=-1, workgroupMappingValue=-1)GEMMSolution(mac_m=64, mac_n=64, mac_k=64, wave_m=32, wave_n=32, wave_k=2, wave_b=1, workgroup_size_x=128, workgroup_size_y=2, workgroupRemapXCC=False, workgroupRemapXCCValue=-1, unroll_x=0, unroll_y=0, load_A='BufferToLDSViaVGPR', load_B='BufferToLDSViaVGPR', storeLDS_D=True, betaInFma=True, scheduler='Priority', schedulerCost='LinearWeighted', prefetch=False, prefetchInFlight=2, prefetchLDSFactor=0, prefetchMixMemOps=False, loadScale_A='BufferToVGPR', loadScale_B='BufferToVGPR', swizzleScale=False, swizzleTileSize={'m': 0, 'k': 0, 'n': 0, 'l': 0}, prefetchScale=False, streamK=True, numWGs=256, streamKTwoTile=True, streamKTwoTileDPFirst=False, architecture={'ArchString': 'gfx950', 'Xnack': False, 'Sramecc': True}, matchMemoryAccess=True, version='')
- SGPR: 72 -> 75 (+3)
	| 3. FloatsGEMM(M: 7680, N: 8448, K: 8192, alpha: 2, beta: 0.5, types: {'type_A': 'half', 'type_B': 'half', 'type_C': 'half', 'type_D': 'half', 'type_acc': 'float', 'trans_A': 'N', 'trans_B': 'T', 'scale_A': 'None', 'scaleType_A': 'None', 'scale_B': 'None', 'scaleType_B': 'None', 'scaleBlockSize': -1, 'scaleSkipPermlane': False, 'scaleShuffleTileA': [], 'scaleShuffleTileB': []}, scaleValue_A: 1, scaleValue_B: 1, initMode_A: DataInitMode(Bounded), initMode_B: DataInitMode(Bounded), initMode_C: DataInitMode(Bounded), workgroupMappingDim: -1, workgroupMappingValue: -1)| GEMM(mac_m: 128, mac_n: 256, mac_k: 16, wave_m: 32, wave_n: 32, wave_k: 8, wave_b: 1, workgroup_size_x: 128, workgroup_size_y: 2, workgroupRemapXCC: False, workgroupRemapXCCValue: -1, unroll_x: 0, unroll_y: 0, load_A: BufferToLDSViaVGPR, load_B: BufferToLDSViaVGPR, storeLDS_D: True, betaInFma: True, scheduler: Priority, schedulerCost: LinearWeighted, prefetch: False, prefetchInFlight: 2, prefetchLDSFactor: 0, prefetchMixMemOps: False, loadScale_A: BufferToVGPR, loadScale_B: BufferToVGPR, swizzleScale: False, swizzleTileSize: {'m': 0, 'k': 0, 'n': 0, 'l': 0}, prefetchScale: False, streamK: True, numWGs: 256, streamKTwoTile: False, streamKTwoTileDPFirst: False, architecture: {'ArchString': 'gfx950', 'Xnack': False, 'Sramecc': True}, matchMemoryAccess: True, version: ) | GEMMProblem(M=7680, N=8448, K=8192, alpha=2, beta=0.5, types={'type_A': 'half', 'type_B': 'half', 'type_C': 'half', 'type_D': 'half', 'type_acc': 'float', 'trans_A': 'N', 'trans_B': 'T', 'scale_A': 'None', 'scaleType_A': 'None', 'scale_B': 'None', 'scaleType_B': 'None', 'scaleBlockSize': -1, 'scaleSkipPermlane': False, 'scaleShuffleTileA': [], 'scaleShuffleTileB': []}, scaleValue_A=1, scaleValue_B=1, initMode_A='DataInitMode(Bounded)', initMode_B='DataInitMode(Bounded)', initMode_C='DataInitMode(Bounded)', workgroupMappingDim=-1, workgroupMappingValue=-1)GEMMSolution(mac_m=128, mac_n=256, mac_k=16, wave_m=32, wave_n=32, wave_k=8, wave_b=1, workgroup_size_x=128, workgroup_size_y=2, workgroupRemapXCC=False, workgroupRemapXCCValue=-1, unroll_x=0, unroll_y=0, load_A='BufferToLDSViaVGPR', load_B='BufferToLDSViaVGPR', storeLDS_D=True, betaInFma=True, scheduler='Priority', schedulerCost='LinearWeighted', prefetch=False, prefetchInFlight=2, prefetchLDSFactor=0, prefetchMixMemOps=False, loadScale_A='BufferToVGPR', loadScale_B='BufferToVGPR', swizzleScale=False, swizzleTileSize={'m': 0, 'k': 0, 'n': 0, 'l': 0}, prefetchScale=False, streamK=True, numWGs=256, streamKTwoTile=False, streamKTwoTileDPFirst=False, architecture={'ArchString': 'gfx950', 'Xnack': False, 'Sramecc': True}, matchMemoryAccess=True, version='')
- SGPR: 73 -> 75 (+2) | VGPR: 154 -> 152 (-2)
	| 3. FloatsGEMM(M: 7680, N: 8448, K: 8192, alpha: 2, beta: 0.5, types: {'type_A': 'half', 'type_B': 'half', 'type_C': 'half', 'type_D': 'half', 'type_acc': 'float', 'trans_A': 'N', 'trans_B': 'T', 'scale_A': 'None', 'scaleType_A': 'None', 'scale_B': 'None', 'scaleType_B': 'None', 'scaleBlockSize': -1, 'scaleSkipPermlane': False, 'scaleShuffleTileA': [], 'scaleShuffleTileB': []}, scaleValue_A: 1, scaleValue_B: 1, initMode_A: DataInitMode(Bounded), initMode_B: DataInitMode(Bounded), initMode_C: DataInitMode(Bounded), workgroupMappingDim: -1, workgroupMappingValue: -1)| GEMM(mac_m: 128, mac_n: 256, mac_k: 16, wave_m: 32, wave_n: 32, wave_k: 8, wave_b: 1, workgroup_size_x: 128, workgroup_size_y: 2, workgroupRemapXCC: False, workgroupRemapXCCValue: -1, unroll_x: 0, unroll_y: 0, load_A: BufferToLDSViaVGPR, load_B: BufferToLDSViaVGPR, storeLDS_D: True, betaInFma: True, scheduler: Priority, schedulerCost: LinearWeighted, prefetch: False, prefetchInFlight: 2, prefetchLDSFactor: 0, prefetchMixMemOps: False, loadScale_A: BufferToVGPR, loadScale_B: BufferToVGPR, swizzleScale: False, swizzleTileSize: {'m': 0, 'k': 0, 'n': 0, 'l': 0}, prefetchScale: False, streamK: True, numWGs: 256, streamKTwoTile: False, streamKTwoTileDPFirst: True, architecture: {'ArchString': 'gfx950', 'Xnack': False, 'Sramecc': True}, matchMemoryAccess: True, version: ) | GEMMProblem(M=7680, N=8448, K=8192, alpha=2, beta=0.5, types={'type_A': 'half', 'type_B': 'half', 'type_C': 'half', 'type_D': 'half', 'type_acc': 'float', 'trans_A': 'N', 'trans_B': 'T', 'scale_A': 'None', 'scaleType_A': 'None', 'scale_B': 'None', 'scaleType_B': 'None', 'scaleBlockSize': -1, 'scaleSkipPermlane': False, 'scaleShuffleTileA': [], 'scaleShuffleTileB': []}, scaleValue_A=1, scaleValue_B=1, initMode_A='DataInitMode(Bounded)', initMode_B='DataInitMode(Bounded)', initMode_C='DataInitMode(Bounded)', workgroupMappingDim=-1, workgroupMappingValue=-1)GEMMSolution(mac_m=128, mac_n=256, mac_k=16, wave_m=32, wave_n=32, wave_k=8, wave_b=1, workgroup_size_x=128, workgroup_size_y=2, workgroupRemapXCC=False, workgroupRemapXCCValue=-1, unroll_x=0, unroll_y=0, load_A='BufferToLDSViaVGPR', load_B='BufferToLDSViaVGPR', storeLDS_D=True, betaInFma=True, scheduler='Priority', schedulerCost='LinearWeighted', prefetch=False, prefetchInFlight=2, prefetchLDSFactor=0, prefetchMixMemOps=False, loadScale_A='BufferToVGPR', loadScale_B='BufferToVGPR', swizzleScale=False, swizzleTileSize={'m': 0, 'k': 0, 'n': 0, 'l': 0}, prefetchScale=False, streamK=True, numWGs=256, streamKTwoTile=False, streamKTwoTileDPFirst=True, architecture={'ArchString': 'gfx950', 'Xnack': False, 'Sramecc': True}, matchMemoryAccess=True, version='')
- SGPR: 75 -> 76 (+1)
	| 3. FloatsGEMM(M: 7680, N: 8448, K: 8192, alpha: 2, beta: 0.5, types: {'type_A': 'half', 'type_B': 'half', 'type_C': 'half', 'type_D': 'half', 'type_acc': 'float', 'trans_A': 'N', 'trans_B': 'T', 'scale_A': 'None', 'scaleType_A': 'None', 'scale_B': 'None', 'scaleType_B': 'None', 'scaleBlockSize': -1, 'scaleSkipPermlane': False, 'scaleShuffleTileA': [], 'scaleShuffleTileB': []}, scaleValue_A: 1, scaleValue_B: 1, initMode_A: DataInitMode(Bounded), initMode_B: DataInitMode(Bounded), initMode_C: DataInitMode(Bounded), workgroupMappingDim: -1, workgroupMappingValue: -1)| GEMM(mac_m: 128, mac_n: 256, mac_k: 16, wave_m: 32, wave_n: 32, wave_k: 8, wave_b: 1, workgroup_size_x: 128, workgroup_size_y: 2, workgroupRemapXCC: False, workgroupRemapXCCValue: -1, unroll_x: 0, unroll_y: 0, load_A: BufferToLDSViaVGPR, load_B: BufferToLDSViaVGPR, storeLDS_D: True, betaInFma: True, scheduler: Priority, schedulerCost: LinearWeighted, prefetch: False, prefetchInFlight: 2, prefetchLDSFactor: 0, prefetchMixMemOps: False, loadScale_A: BufferToVGPR, loadScale_B: BufferToVGPR, swizzleScale: False, swizzleTileSize: {'m': 0, 'k': 0, 'n': 0, 'l': 0}, prefetchScale: False, streamK: True, numWGs: 256, streamKTwoTile: True, streamKTwoTileDPFirst: False, architecture: {'ArchString': 'gfx950', 'Xnack': False, 'Sramecc': True}, matchMemoryAccess: True, version: ) | GEMMProblem(M=7680, N=8448, K=8192, alpha=2, beta=0.5, types={'type_A': 'half', 'type_B': 'half', 'type_C': 'half', 'type_D': 'half', 'type_acc': 'float', 'trans_A': 'N', 'trans_B': 'T', 'scale_A': 'None', 'scaleType_A': 'None', 'scale_B': 'None', 'scaleType_B': 'None', 'scaleBlockSize': -1, 'scaleSkipPermlane': False, 'scaleShuffleTileA': [], 'scaleShuffleTileB': []}, scaleValue_A=1, scaleValue_B=1, initMode_A='DataInitMode(Bounded)', initMode_B='DataInitMode(Bounded)', initMode_C='DataInitMode(Bounded)', workgroupMappingDim=-1, workgroupMappingValue=-1)GEMMSolution(mac_m=128, mac_n=256, mac_k=16, wave_m=32, wave_n=32, wave_k=8, wave_b=1, workgroup_size_x=128, workgroup_size_y=2, workgroupRemapXCC=False, workgroupRemapXCCValue=-1, unroll_x=0, unroll_y=0, load_A='BufferToLDSViaVGPR', load_B='BufferToLDSViaVGPR', storeLDS_D=True, betaInFma=True, scheduler='Priority', schedulerCost='LinearWeighted', prefetch=False, prefetchInFlight=2, prefetchLDSFactor=0, prefetchMixMemOps=False, loadScale_A='BufferToVGPR', loadScale_B='BufferToVGPR', swizzleScale=False, swizzleTileSize={'m': 0, 'k': 0, 'n': 0, 'l': 0}, prefetchScale=False, streamK=True, numWGs=256, streamKTwoTile=True, streamKTwoTileDPFirst=False, architecture={'ArchString': 'gfx950', 'Xnack': False, 'Sramecc': True}, matchMemoryAccess=True, version='')
- SGPR: 72 -> 75 (+3)
	| 3. FloatsGEMM(M: 7680, N: 8448, K: 8448, alpha: 2, beta: 0.5, types: {'type_A': 'half', 'type_B': 'half', 'type_C': 'half', 'type_D': 'half', 'type_acc': 'float', 'trans_A': 'N', 'trans_B': 'T', 'scale_A': 'None', 'scaleType_A': 'None', 'scale_B': 'None', 'scaleType_B': 'None', 'scaleBlockSize': -1, 'scaleSkipPermlane': False, 'scaleShuffleTileA': [], 'scaleShuffleTileB': []}, scaleValue_A: 1, scaleValue_B: 1, initMode_A: DataInitMode(Bounded), initMode_B: DataInitMode(Bounded), initMode_C: DataInitMode(Bounded), workgroupMappingDim: -1, workgroupMappingValue: -1)| GEMM(mac_m: 128, mac_n: 128, mac_k: 16, wave_m: 32, wave_n: 32, wave_k: 8, wave_b: 1, workgroup_size_x: 128, workgroup_size_y: 2, workgroupRemapXCC: False, workgroupRemapXCCValue: -1, unroll_x: 0, unroll_y: 0, load_A: BufferToLDSViaVGPR, load_B: BufferToLDSViaVGPR, storeLDS_D: True, betaInFma: True, scheduler: Priority, schedulerCost: LinearWeighted, prefetch: False, prefetchInFlight: 0, prefetchLDSFactor: 0, prefetchMixMemOps: False, loadScale_A: BufferToVGPR, loadScale_B: BufferToVGPR, swizzleScale: False, swizzleTileSize: {'m': 0, 'k': 0, 'n': 0, 'l': 0}, prefetchScale: False, streamK: True, numWGs: 256, streamKTwoTile: False, streamKTwoTileDPFirst: False, architecture: {'ArchString': 'gfx950', 'Xnack': False, 'Sramecc': True}, matchMemoryAccess: True, version: ) | GEMMProblem(M=7680, N=8448, K=8448, alpha=2, beta=0.5, types={'type_A': 'half', 'type_B': 'half', 'type_C': 'half', 'type_D': 'half', 'type_acc': 'float', 'trans_A': 'N', 'trans_B': 'T', 'scale_A': 'None', 'scaleType_A': 'None', 'scale_B': 'None', 'scaleType_B': 'None', 'scaleBlockSize': -1, 'scaleSkipPermlane': False, 'scaleShuffleTileA': [], 'scaleShuffleTileB': []}, scaleValue_A=1, scaleValue_B=1, initMode_A='DataInitMode(Bounded)', initMode_B='DataInitMode(Bounded)', initMode_C='DataInitMode(Bounded)', workgroupMappingDim=-1, workgroupMappingValue=-1)GEMMSolution(mac_m=128, mac_n=128, mac_k=16, wave_m=32, wave_n=32, wave_k=8, wave_b=1, workgroup_size_x=128, workgroup_size_y=2, workgroupRemapXCC=False, workgroupRemapXCCValue=-1, unroll_x=0, unroll_y=0, load_A='BufferToLDSViaVGPR', load_B='BufferToLDSViaVGPR', storeLDS_D=True, betaInFma=True, scheduler='Priority', schedulerCost='LinearWeighted', prefetch=False, prefetchInFlight=0, prefetchLDSFactor=0, prefetchMixMemOps=False, loadScale_A='BufferToVGPR', loadScale_B='BufferToVGPR', swizzleScale=False, swizzleTileSize={'m': 0, 'k': 0, 'n': 0, 'l': 0}, prefetchScale=False, streamK=True, numWGs=256, streamKTwoTile=False, streamKTwoTileDPFirst=False, architecture={'ArchString': 'gfx950', 'Xnack': False, 'Sramecc': True}, matchMemoryAccess=True, version='')
- SGPR: 73 -> 75 (+2)
	| 3. FloatsGEMM(M: 7680, N: 8448, K: 8448, alpha: 2, beta: 0.5, types: {'type_A': 'half', 'type_B': 'half', 'type_C': 'half', 'type_D': 'half', 'type_acc': 'float', 'trans_A': 'N', 'trans_B': 'T', 'scale_A': 'None', 'scaleType_A': 'None', 'scale_B': 'None', 'scaleType_B': 'None', 'scaleBlockSize': -1, 'scaleSkipPermlane': False, 'scaleShuffleTileA': [], 'scaleShuffleTileB': []}, scaleValue_A: 1, scaleValue_B: 1, initMode_A: DataInitMode(Bounded), initMode_B: DataInitMode(Bounded), initMode_C: DataInitMode(Bounded), workgroupMappingDim: -1, workgroupMappingValue: -1)| GEMM(mac_m: 128, mac_n: 128, mac_k: 16, wave_m: 32, wave_n: 32, wave_k: 8, wave_b: 1, workgroup_size_x: 128, workgroup_size_y: 2, workgroupRemapXCC: False, workgroupRemapXCCValue: -1, unroll_x: 0, unroll_y: 0, load_A: BufferToLDSViaVGPR, load_B: BufferToLDSViaVGPR, storeLDS_D: True, betaInFma: True, scheduler: Priority, schedulerCost: LinearWeighted, prefetch: False, prefetchInFlight: 0, prefetchLDSFactor: 0, prefetchMixMemOps: False, loadScale_A: BufferToVGPR, loadScale_B: BufferToVGPR, swizzleScale: False, swizzleTileSize: {'m': 0, 'k': 0, 'n': 0, 'l': 0}, prefetchScale: False, streamK: True, numWGs: 256, streamKTwoTile: False, streamKTwoTileDPFirst: True, architecture: {'ArchString': 'gfx950', 'Xnack': False, 'Sramecc': True}, matchMemoryAccess: True, version: ) | GEMMProblem(M=7680, N=8448, K=8448, alpha=2, beta=0.5, types={'type_A': 'half', 'type_B': 'half', 'type_C': 'half', 'type_D': 'half', 'type_acc': 'float', 'trans_A': 'N', 'trans_B': 'T', 'scale_A': 'None', 'scaleType_A': 'None', 'scale_B': 'None', 'scaleType_B': 'None', 'scaleBlockSize': -1, 'scaleSkipPermlane': False, 'scaleShuffleTileA': [], 'scaleShuffleTileB': []}, scaleValue_A=1, scaleValue_B=1, initMode_A='DataInitMode(Bounded)', initMode_B='DataInitMode(Bounded)', initMode_C='DataInitMode(Bounded)', workgroupMappingDim=-1, workgroupMappingValue=-1)GEMMSolution(mac_m=128, mac_n=128, mac_k=16, wave_m=32, wave_n=32, wave_k=8, wave_b=1, workgroup_size_x=128, workgroup_size_y=2, workgroupRemapXCC=False, workgroupRemapXCCValue=-1, unroll_x=0, unroll_y=0, load_A='BufferToLDSViaVGPR', load_B='BufferToLDSViaVGPR', storeLDS_D=True, betaInFma=True, scheduler='Priority', schedulerCost='LinearWeighted', prefetch=False, prefetchInFlight=0, prefetchLDSFactor=0, prefetchMixMemOps=False, loadScale_A='BufferToVGPR', loadScale_B='BufferToVGPR', swizzleScale=False, swizzleTileSize={'m': 0, 'k': 0, 'n': 0, 'l': 0}, prefetchScale=False, streamK=True, numWGs=256, streamKTwoTile=False, streamKTwoTileDPFirst=True, architecture={'ArchString': 'gfx950', 'Xnack': False, 'Sramecc': True}, matchMemoryAccess=True, version='')
- SGPR: 75 -> 76 (+1)
	| 3. FloatsGEMM(M: 7680, N: 8448, K: 8448, alpha: 2, beta: 0.5, types: {'type_A': 'half', 'type_B': 'half', 'type_C': 'half', 'type_D': 'half', 'type_acc': 'float', 'trans_A': 'N', 'trans_B': 'T', 'scale_A': 'None', 'scaleType_A': 'None', 'scale_B': 'None', 'scaleType_B': 'None', 'scaleBlockSize': -1, 'scaleSkipPermlane': False, 'scaleShuffleTileA': [], 'scaleShuffleTileB': []}, scaleValue_A: 1, scaleValue_B: 1, initMode_A: DataInitMode(Bounded), initMode_B: DataInitMode(Bounded), initMode_C: DataInitMode(Bounded), workgroupMappingDim: -1, workgroupMappingValue: -1)| GEMM(mac_m: 128, mac_n: 128, mac_k: 16, wave_m: 32, wave_n: 32, wave_k: 8, wave_b: 1, workgroup_size_x: 128, workgroup_size_y: 2, workgroupRemapXCC: False, workgroupRemapXCCValue: -1, unroll_x: 0, unroll_y: 0, load_A: BufferToLDSViaVGPR, load_B: BufferToLDSViaVGPR, storeLDS_D: True, betaInFma: True, scheduler: Priority, schedulerCost: LinearWeighted, prefetch: False, prefetchInFlight: 0, prefetchLDSFactor: 0, prefetchMixMemOps: False, loadScale_A: BufferToVGPR, loadScale_B: BufferToVGPR, swizzleScale: False, swizzleTileSize: {'m': 0, 'k': 0, 'n': 0, 'l': 0}, prefetchScale: False, streamK: True, numWGs: 256, streamKTwoTile: True, streamKTwoTileDPFirst: False, architecture: {'ArchString': 'gfx950', 'Xnack': False, 'Sramecc': True}, matchMemoryAccess: True, version: ) | GEMMProblem(M=7680, N=8448, K=8448, alpha=2, beta=0.5, types={'type_A': 'half', 'type_B': 'half', 'type_C': 'half', 'type_D': 'half', 'type_acc': 'float', 'trans_A': 'N', 'trans_B': 'T', 'scale_A': 'None', 'scaleType_A': 'None', 'scale_B': 'None', 'scaleType_B': 'None', 'scaleBlockSize': -1, 'scaleSkipPermlane': False, 'scaleShuffleTileA': [], 'scaleShuffleTileB': []}, scaleValue_A=1, scaleValue_B=1, initMode_A='DataInitMode(Bounded)', initMode_B='DataInitMode(Bounded)', initMode_C='DataInitMode(Bounded)', workgroupMappingDim=-1, workgroupMappingValue=-1)GEMMSolution(mac_m=128, mac_n=128, mac_k=16, wave_m=32, wave_n=32, wave_k=8, wave_b=1, workgroup_size_x=128, workgroup_size_y=2, workgroupRemapXCC=False, workgroupRemapXCCValue=-1, unroll_x=0, unroll_y=0, load_A='BufferToLDSViaVGPR', load_B='BufferToLDSViaVGPR', storeLDS_D=True, betaInFma=True, scheduler='Priority', schedulerCost='LinearWeighted', prefetch=False, prefetchInFlight=0, prefetchLDSFactor=0, prefetchMixMemOps=False, loadScale_A='BufferToVGPR', loadScale_B='BufferToVGPR', swizzleScale=False, swizzleTileSize={'m': 0, 'k': 0, 'n': 0, 'l': 0}, prefetchScale=False, streamK=True, numWGs=256, streamKTwoTile=True, streamKTwoTileDPFirst=False, architecture={'ArchString': 'gfx950', 'Xnack': False, 'Sramecc': True}, matchMemoryAccess=True, version='')
- SGPR: 72 -> 75 (+3)
	| 3. FloatsGEMM(M: 7680, N: 8448, K: 8448, alpha: 2, beta: 0.5, types: {'type_A': 'half', 'type_B': 'half', 'type_C': 'half', 'type_D': 'half', 'type_acc': 'float', 'trans_A': 'N', 'trans_B': 'T', 'scale_A': 'None', 'scaleType_A': 'None', 'scale_B': 'None', 'scaleType_B': 'None', 'scaleBlockSize': -1, 'scaleSkipPermlane': False, 'scaleShuffleTileA': [], 'scaleShuffleTileB': []}, scaleValue_A: 1, scaleValue_B: 1, initMode_A: DataInitMode(Bounded), initMode_B: DataInitMode(Bounded), initMode_C: DataInitMode(Bounded), workgroupMappingDim: -1, workgroupMappingValue: -1)| GEMM(mac_m: 128, mac_n: 128, mac_k: 16, wave_m: 32, wave_n: 32, wave_k: 8, wave_b: 1, workgroup_size_x: 128, workgroup_size_y: 2, workgroupRemapXCC: False, workgroupRemapXCCValue: -1, unroll_x: 0, unroll_y: 0, load_A: BufferToLDSViaVGPR, load_B: BufferToLDSViaVGPR, storeLDS_D: True, betaInFma: True, scheduler: Priority, schedulerCost: LinearWeighted, prefetch: True, prefetchInFlight: 2, prefetchLDSFactor: 2, prefetchMixMemOps: False, loadScale_A: BufferToVGPR, loadScale_B: BufferToVGPR, swizzleScale: False, swizzleTileSize: {'m': 0, 'k': 0, 'n': 0, 'l': 0}, prefetchScale: False, streamK: True, numWGs: 256, streamKTwoTile: False, streamKTwoTileDPFirst: False, architecture: {'ArchString': 'gfx950', 'Xnack': False, 'Sramecc': True}, matchMemoryAccess: True, version: ) | GEMMProblem(M=7680, N=8448, K=8448, alpha=2, beta=0.5, types={'type_A': 'half', 'type_B': 'half', 'type_C': 'half', 'type_D': 'half', 'type_acc': 'float', 'trans_A': 'N', 'trans_B': 'T', 'scale_A': 'None', 'scaleType_A': 'None', 'scale_B': 'None', 'scaleType_B': 'None', 'scaleBlockSize': -1, 'scaleSkipPermlane': False, 'scaleShuffleTileA': [], 'scaleShuffleTileB': []}, scaleValue_A=1, scaleValue_B=1, initMode_A='DataInitMode(Bounded)', initMode_B='DataInitMode(Bounded)', initMode_C='DataInitMode(Bounded)', workgroupMappingDim=-1, workgroupMappingValue=-1)GEMMSolution(mac_m=128, mac_n=128, mac_k=16, wave_m=32, wave_n=32, wave_k=8, wave_b=1, workgroup_size_x=128, workgroup_size_y=2, workgroupRemapXCC=False, workgroupRemapXCCValue=-1, unroll_x=0, unroll_y=0, load_A='BufferToLDSViaVGPR', load_B='BufferToLDSViaVGPR', storeLDS_D=True, betaInFma=True, scheduler='Priority', schedulerCost='LinearWeighted', prefetch=True, prefetchInFlight=2, prefetchLDSFactor=2, prefetchMixMemOps=False, loadScale_A='BufferToVGPR', loadScale_B='BufferToVGPR', swizzleScale=False, swizzleTileSize={'m': 0, 'k': 0, 'n': 0, 'l': 0}, prefetchScale=False, streamK=True, numWGs=256, streamKTwoTile=False, streamKTwoTileDPFirst=False, architecture={'ArchString': 'gfx950', 'Xnack': False, 'Sramecc': True}, matchMemoryAccess=True, version='')
- SGPR: 72 -> 75 (+3)
	| 3. FloatsGEMM(M: 7680, N: 8448, K: 8448, alpha: 2, beta: 0.5, types: {'type_A': 'half', 'type_B': 'half', 'type_C': 'half', 'type_D': 'half', 'type_acc': 'float', 'trans_A': 'N', 'trans_B': 'T', 'scale_A': 'None', 'scaleType_A': 'None', 'scale_B': 'None', 'scaleType_B': 'None', 'scaleBlockSize': -1, 'scaleSkipPermlane': False, 'scaleShuffleTileA': [], 'scaleShuffleTileB': []}, scaleValue_A: 1, scaleValue_B: 1, initMode_A: DataInitMode(Bounded), initMode_B: DataInitMode(Bounded), initMode_C: DataInitMode(Bounded), workgroupMappingDim: -1, workgroupMappingValue: -1)| GEMM(mac_m: 128, mac_n: 128, mac_k: 32, wave_m: 32, wave_n: 32, wave_k: 8, wave_b: 1, workgroup_size_x: 128, workgroup_size_y: 2, workgroupRemapXCC: False, workgroupRemapXCCValue: -1, unroll_x: 0, unroll_y: 0, load_A: BufferToLDSViaVGPR, load_B: BufferToLDSViaVGPR, storeLDS_D: True, betaInFma: True, scheduler: Priority, schedulerCost: LinearWeighted, prefetch: False, prefetchInFlight: 0, prefetchLDSFactor: 0, prefetchMixMemOps: False, loadScale_A: BufferToVGPR, loadScale_B: BufferToVGPR, swizzleScale: False, swizzleTileSize: {'m': 0, 'k': 0, 'n': 0, 'l': 0}, prefetchScale: False, streamK: True, numWGs: 256, streamKTwoTile: False, streamKTwoTileDPFirst: False, architecture: {'ArchString': 'gfx950', 'Xnack': False, 'Sramecc': True}, matchMemoryAccess: True, version: ) | GEMMProblem(M=7680, N=8448, K=8448, alpha=2, beta=0.5, types={'type_A': 'half', 'type_B': 'half', 'type_C': 'half', 'type_D': 'half', 'type_acc': 'float', 'trans_A': 'N', 'trans_B': 'T', 'scale_A': 'None', 'scaleType_A': 'None', 'scale_B': 'None', 'scaleType_B': 'None', 'scaleBlockSize': -1, 'scaleSkipPermlane': False, 'scaleShuffleTileA': [], 'scaleShuffleTileB': []}, scaleValue_A=1, scaleValue_B=1, initMode_A='DataInitMode(Bounded)', initMode_B='DataInitMode(Bounded)', initMode_C='DataInitMode(Bounded)', workgroupMappingDim=-1, workgroupMappingValue=-1)GEMMSolution(mac_m=128, mac_n=128, mac_k=32, wave_m=32, wave_n=32, wave_k=8, wave_b=1, workgroup_size_x=128, workgroup_size_y=2, workgroupRemapXCC=False, workgroupRemapXCCValue=-1, unroll_x=0, unroll_y=0, load_A='BufferToLDSViaVGPR', load_B='BufferToLDSViaVGPR', storeLDS_D=True, betaInFma=True, scheduler='Priority', schedulerCost='LinearWeighted', prefetch=False, prefetchInFlight=0, prefetchLDSFactor=0, prefetchMixMemOps=False, loadScale_A='BufferToVGPR', loadScale_B='BufferToVGPR', swizzleScale=False, swizzleTileSize={'m': 0, 'k': 0, 'n': 0, 'l': 0}, prefetchScale=False, streamK=True, numWGs=256, streamKTwoTile=False, streamKTwoTileDPFirst=False, architecture={'ArchString': 'gfx950', 'Xnack': False, 'Sramecc': True}, matchMemoryAccess=True, version='')
- SGPR: 73 -> 75 (+2) | VGPR: 182 -> 180 (-2)
	| 3. FloatsGEMM(M: 7680, N: 8448, K: 8448, alpha: 2, beta: 0.5, types: {'type_A': 'half', 'type_B': 'half', 'type_C': 'half', 'type_D': 'half', 'type_acc': 'float', 'trans_A': 'N', 'trans_B': 'T', 'scale_A': 'None', 'scaleType_A': 'None', 'scale_B': 'None', 'scaleType_B': 'None', 'scaleBlockSize': -1, 'scaleSkipPermlane': False, 'scaleShuffleTileA': [], 'scaleShuffleTileB': []}, scaleValue_A: 1, scaleValue_B: 1, initMode_A: DataInitMode(Bounded), initMode_B: DataInitMode(Bounded), initMode_C: DataInitMode(Bounded), workgroupMappingDim: -1, workgroupMappingValue: -1)| GEMM(mac_m: 128, mac_n: 128, mac_k: 32, wave_m: 32, wave_n: 32, wave_k: 8, wave_b: 1, workgroup_size_x: 128, workgroup_size_y: 2, workgroupRemapXCC: False, workgroupRemapXCCValue: -1, unroll_x: 0, unroll_y: 0, load_A: BufferToLDSViaVGPR, load_B: BufferToLDSViaVGPR, storeLDS_D: True, betaInFma: True, scheduler: Priority, schedulerCost: LinearWeighted, prefetch: False, prefetchInFlight: 0, prefetchLDSFactor: 0, prefetchMixMemOps: False, loadScale_A: BufferToVGPR, loadScale_B: BufferToVGPR, swizzleScale: False, swizzleTileSize: {'m': 0, 'k': 0, 'n': 0, 'l': 0}, prefetchScale: False, streamK: True, numWGs: 256, streamKTwoTile: False, streamKTwoTileDPFirst: True, architecture: {'ArchString': 'gfx950', 'Xnack': False, 'Sramecc': True}, matchMemoryAccess: True, version: ) | GEMMProblem(M=7680, N=8448, K=8448, alpha=2, beta=0.5, types={'type_A': 'half', 'type_B': 'half', 'type_C': 'half', 'type_D': 'half', 'type_acc': 'float', 'trans_A': 'N', 'trans_B': 'T', 'scale_A': 'None', 'scaleType_A': 'None', 'scale_B': 'None', 'scaleType_B': 'None', 'scaleBlockSize': -1, 'scaleSkipPermlane': False, 'scaleShuffleTileA': [], 'scaleShuffleTileB': []}, scaleValue_A=1, scaleValue_B=1, initMode_A='DataInitMode(Bounded)', initMode_B='DataInitMode(Bounded)', initMode_C='DataInitMode(Bounded)', workgroupMappingDim=-1, workgroupMappingValue=-1)GEMMSolution(mac_m=128, mac_n=128, mac_k=32, wave_m=32, wave_n=32, wave_k=8, wave_b=1, workgroup_size_x=128, workgroup_size_y=2, workgroupRemapXCC=False, workgroupRemapXCCValue=-1, unroll_x=0, unroll_y=0, load_A='BufferToLDSViaVGPR', load_B='BufferToLDSViaVGPR', storeLDS_D=True, betaInFma=True, scheduler='Priority', schedulerCost='LinearWeighted', prefetch=False, prefetchInFlight=0, prefetchLDSFactor=0, prefetchMixMemOps=False, loadScale_A='BufferToVGPR', loadScale_B='BufferToVGPR', swizzleScale=False, swizzleTileSize={'m': 0, 'k': 0, 'n': 0, 'l': 0}, prefetchScale=False, streamK=True, numWGs=256, streamKTwoTile=False, streamKTwoTileDPFirst=True, architecture={'ArchString': 'gfx950', 'Xnack': False, 'Sramecc': True}, matchMemoryAccess=True, version='')
- SGPR: 75 -> 76 (+1)
	| 3. FloatsGEMM(M: 7680, N: 8448, K: 8448, alpha: 2, beta: 0.5, types: {'type_A': 'half', 'type_B': 'half', 'type_C': 'half', 'type_D': 'half', 'type_acc': 'float', 'trans_A': 'N', 'trans_B': 'T', 'scale_A': 'None', 'scaleType_A': 'None', 'scale_B': 'None', 'scaleType_B': 'None', 'scaleBlockSize': -1, 'scaleSkipPermlane': False, 'scaleShuffleTileA': [], 'scaleShuffleTileB': []}, scaleValue_A: 1, scaleValue_B: 1, initMode_A: DataInitMode(Bounded), initMode_B: DataInitMode(Bounded), initMode_C: DataInitMode(Bounded), workgroupMappingDim: -1, workgroupMappingValue: -1)| GEMM(mac_m: 128, mac_n: 128, mac_k: 32, wave_m: 32, wave_n: 32, wave_k: 8, wave_b: 1, workgroup_size_x: 128, workgroup_size_y: 2, workgroupRemapXCC: False, workgroupRemapXCCValue: -1, unroll_x: 0, unroll_y: 0, load_A: BufferToLDSViaVGPR, load_B: BufferToLDSViaVGPR, storeLDS_D: True, betaInFma: True, scheduler: Priority, schedulerCost: LinearWeighted, prefetch: False, prefetchInFlight: 0, prefetchLDSFactor: 0, prefetchMixMemOps: False, loadScale_A: BufferToVGPR, loadScale_B: BufferToVGPR, swizzleScale: False, swizzleTileSize: {'m': 0, 'k': 0, 'n': 0, 'l': 0}, prefetchScale: False, streamK: True, numWGs: 256, streamKTwoTile: True, streamKTwoTileDPFirst: False, architecture: {'ArchString': 'gfx950', 'Xnack': False, 'Sramecc': True}, matchMemoryAccess: True, version: ) | GEMMProblem(M=7680, N=8448, K=8448, alpha=2, beta=0.5, types={'type_A': 'half', 'type_B': 'half', 'type_C': 'half', 'type_D': 'half', 'type_acc': 'float', 'trans_A': 'N', 'trans_B': 'T', 'scale_A': 'None', 'scaleType_A': 'None', 'scale_B': 'None', 'scaleType_B': 'None', 'scaleBlockSize': -1, 'scaleSkipPermlane': False, 'scaleShuffleTileA': [], 'scaleShuffleTileB': []}, scaleValue_A=1, scaleValue_B=1, initMode_A='DataInitMode(Bounded)', initMode_B='DataInitMode(Bounded)', initMode_C='DataInitMode(Bounded)', workgroupMappingDim=-1, workgroupMappingValue=-1)GEMMSolution(mac_m=128, mac_n=128, mac_k=32, wave_m=32, wave_n=32, wave_k=8, wave_b=1, workgroup_size_x=128, workgroup_size_y=2, workgroupRemapXCC=False, workgroupRemapXCCValue=-1, unroll_x=0, unroll_y=0, load_A='BufferToLDSViaVGPR', load_B='BufferToLDSViaVGPR', storeLDS_D=True, betaInFma=True, scheduler='Priority', schedulerCost='LinearWeighted', prefetch=False, prefetchInFlight=0, prefetchLDSFactor=0, prefetchMixMemOps=False, loadScale_A='BufferToVGPR', loadScale_B='BufferToVGPR', swizzleScale=False, swizzleTileSize={'m': 0, 'k': 0, 'n': 0, 'l': 0}, prefetchScale=False, streamK=True, numWGs=256, streamKTwoTile=True, streamKTwoTileDPFirst=False, architecture={'ArchString': 'gfx950', 'Xnack': False, 'Sramecc': True}, matchMemoryAccess=True, version='')
- SGPR: 73 -> 75 (+2)
	| 3. FloatsGEMM(M: 7680, N: 8448, K: 8448, alpha: 2, beta: 0.5, types: {'type_A': 'half', 'type_B': 'half', 'type_C': 'half', 'type_D': 'half', 'type_acc': 'float', 'trans_A': 'N', 'trans_B': 'T', 'scale_A': 'None', 'scaleType_A': 'None', 'scale_B': 'None', 'scaleType_B': 'None', 'scaleBlockSize': -1, 'scaleSkipPermlane': False, 'scaleShuffleTileA': [], 'scaleShuffleTileB': []}, scaleValue_A: 1, scaleValue_B: 1, initMode_A: DataInitMode(Bounded), initMode_B: DataInitMode(Bounded), initMode_C: DataInitMode(Bounded), workgroupMappingDim: -1, workgroupMappingValue: -1)| GEMM(mac_m: 128, mac_n: 128, mac_k: 32, wave_m: 32, wave_n: 32, wave_k: 8, wave_b: 1, workgroup_size_x: 128, workgroup_size_y: 2, workgroupRemapXCC: False, workgroupRemapXCCValue: -1, unroll_x: 0, unroll_y: 0, load_A: BufferToLDSViaVGPR, load_B: BufferToLDSViaVGPR, storeLDS_D: True, betaInFma: True, scheduler: Priority, schedulerCost: LinearWeighted, prefetch: True, prefetchInFlight: 2, prefetchLDSFactor: 2, prefetchMixMemOps: False, loadScale_A: BufferToVGPR, loadScale_B: BufferToVGPR, swizzleScale: False, swizzleTileSize: {'m': 0, 'k': 0, 'n': 0, 'l': 0}, prefetchScale: False, streamK: True, numWGs: 256, streamKTwoTile: False, streamKTwoTileDPFirst: False, architecture: {'ArchString': 'gfx950', 'Xnack': False, 'Sramecc': True}, matchMemoryAccess: True, version: ) | GEMMProblem(M=7680, N=8448, K=8448, alpha=2, beta=0.5, types={'type_A': 'half', 'type_B': 'half', 'type_C': 'half', 'type_D': 'half', 'type_acc': 'float', 'trans_A': 'N', 'trans_B': 'T', 'scale_A': 'None', 'scaleType_A': 'None', 'scale_B': 'None', 'scaleType_B': 'None', 'scaleBlockSize': -1, 'scaleSkipPermlane': False, 'scaleShuffleTileA': [], 'scaleShuffleTileB': []}, scaleValue_A=1, scaleValue_B=1, initMode_A='DataInitMode(Bounded)', initMode_B='DataInitMode(Bounded)', initMode_C='DataInitMode(Bounded)', workgroupMappingDim=-1, workgroupMappingValue=-1)GEMMSolution(mac_m=128, mac_n=128, mac_k=32, wave_m=32, wave_n=32, wave_k=8, wave_b=1, workgroup_size_x=128, workgroup_size_y=2, workgroupRemapXCC=False, workgroupRemapXCCValue=-1, unroll_x=0, unroll_y=0, load_A='BufferToLDSViaVGPR', load_B='BufferToLDSViaVGPR', storeLDS_D=True, betaInFma=True, scheduler='Priority', schedulerCost='LinearWeighted', prefetch=True, prefetchInFlight=2, prefetchLDSFactor=2, prefetchMixMemOps=False, loadScale_A='BufferToVGPR', loadScale_B='BufferToVGPR', swizzleScale=False, swizzleTileSize={'m': 0, 'k': 0, 'n': 0, 'l': 0}, prefetchScale=False, streamK=True, numWGs=256, streamKTwoTile=False, streamKTwoTileDPFirst=False, architecture={'ArchString': 'gfx950', 'Xnack': False, 'Sramecc': True}, matchMemoryAccess=True, version='')
- SGPR: 72 -> 75 (+3)
	| 3. FloatsGEMM(M: 7680, N: 8448, K: 8448, alpha: 2, beta: 0.5, types: {'type_A': 'half', 'type_B': 'half', 'type_C': 'half', 'type_D': 'half', 'type_acc': 'float', 'trans_A': 'N', 'trans_B': 'T', 'scale_A': 'None', 'scaleType_A': 'None', 'scale_B': 'None', 'scaleType_B': 'None', 'scaleBlockSize': -1, 'scaleSkipPermlane': False, 'scaleShuffleTileA': [], 'scaleShuffleTileB': []}, scaleValue_A: 1, scaleValue_B: 1, initMode_A: DataInitMode(Bounded), initMode_B: DataInitMode(Bounded), initMode_C: DataInitMode(Bounded), workgroupMappingDim: -1, workgroupMappingValue: -1)| GEMM(mac_m: 128, mac_n: 128, mac_k: 64, wave_m: 32, wave_n: 32, wave_k: 8, wave_b: 1, workgroup_size_x: 128, workgroup_size_y: 2, workgroupRemapXCC: False, workgroupRemapXCCValue: -1, unroll_x: 0, unroll_y: 0, load_A: BufferToLDSViaVGPR, load_B: BufferToLDSViaVGPR, storeLDS_D: True, betaInFma: True, scheduler: Priority, schedulerCost: LinearWeighted, prefetch: False, prefetchInFlight: 0, prefetchLDSFactor: 0, prefetchMixMemOps: False, loadScale_A: BufferToVGPR, loadScale_B: BufferToVGPR, swizzleScale: False, swizzleTileSize: {'m': 0, 'k': 0, 'n': 0, 'l': 0}, prefetchScale: False, streamK: True, numWGs: 256, streamKTwoTile: False, streamKTwoTileDPFirst: False, architecture: {'ArchString': 'gfx950', 'Xnack': False, 'Sramecc': True}, matchMemoryAccess: True, version: ) | GEMMProblem(M=7680, N=8448, K=8448, alpha=2, beta=0.5, types={'type_A': 'half', 'type_B': 'half', 'type_C': 'half', 'type_D': 'half', 'type_acc': 'float', 'trans_A': 'N', 'trans_B': 'T', 'scale_A': 'None', 'scaleType_A': 'None', 'scale_B': 'None', 'scaleType_B': 'None', 'scaleBlockSize': -1, 'scaleSkipPermlane': False, 'scaleShuffleTileA': [], 'scaleShuffleTileB': []}, scaleValue_A=1, scaleValue_B=1, initMode_A='DataInitMode(Bounded)', initMode_B='DataInitMode(Bounded)', initMode_C='DataInitMode(Bounded)', workgroupMappingDim=-1, workgroupMappingValue=-1)GEMMSolution(mac_m=128, mac_n=128, mac_k=64, wave_m=32, wave_n=32, wave_k=8, wave_b=1, workgroup_size_x=128, workgroup_size_y=2, workgroupRemapXCC=False, workgroupRemapXCCValue=-1, unroll_x=0, unroll_y=0, load_A='BufferToLDSViaVGPR', load_B='BufferToLDSViaVGPR', storeLDS_D=True, betaInFma=True, scheduler='Priority', schedulerCost='LinearWeighted', prefetch=False, prefetchInFlight=0, prefetchLDSFactor=0, prefetchMixMemOps=False, loadScale_A='BufferToVGPR', loadScale_B='BufferToVGPR', swizzleScale=False, swizzleTileSize={'m': 0, 'k': 0, 'n': 0, 'l': 0}, prefetchScale=False, streamK=True, numWGs=256, streamKTwoTile=False, streamKTwoTileDPFirst=False, architecture={'ArchString': 'gfx950', 'Xnack': False, 'Sramecc': True}, matchMemoryAccess=True, version='')
- SGPR: 73 -> 75 (+2)
	| 3. FloatsGEMM(M: 7680, N: 8448, K: 8448, alpha: 2, beta: 0.5, types: {'type_A': 'half', 'type_B': 'half', 'type_C': 'half', 'type_D': 'half', 'type_acc': 'float', 'trans_A': 'N', 'trans_B': 'T', 'scale_A': 'None', 'scaleType_A': 'None', 'scale_B': 'None', 'scaleType_B': 'None', 'scaleBlockSize': -1, 'scaleSkipPermlane': False, 'scaleShuffleTileA': [], 'scaleShuffleTileB': []}, scaleValue_A: 1, scaleValue_B: 1, initMode_A: DataInitMode(Bounded), initMode_B: DataInitMode(Bounded), initMode_C: DataInitMode(Bounded), workgroupMappingDim: -1, workgroupMappingValue: -1)| GEMM(mac_m: 128, mac_n: 128, mac_k: 64, wave_m: 32, wave_n: 32, wave_k: 8, wave_b: 1, workgroup_size_x: 128, workgroup_size_y: 2, workgroupRemapXCC: False, workgroupRemapXCCValue: -1, unroll_x: 0, unroll_y: 0, load_A: BufferToLDSViaVGPR, load_B: BufferToLDSViaVGPR, storeLDS_D: True, betaInFma: True, scheduler: Priority, schedulerCost: LinearWeighted, prefetch: True, prefetchInFlight: 2, prefetchLDSFactor: 2, prefetchMixMemOps: False, loadScale_A: BufferToVGPR, loadScale_B: BufferToVGPR, swizzleScale: False, swizzleTileSize: {'m': 0, 'k': 0, 'n': 0, 'l': 0}, prefetchScale: False, streamK: True, numWGs: 256, streamKTwoTile: False, streamKTwoTileDPFirst: False, architecture: {'ArchString': 'gfx950', 'Xnack': False, 'Sramecc': True}, matchMemoryAccess: True, version: ) | GEMMProblem(M=7680, N=8448, K=8448, alpha=2, beta=0.5, types={'type_A': 'half', 'type_B': 'half', 'type_C': 'half', 'type_D': 'half', 'type_acc': 'float', 'trans_A': 'N', 'trans_B': 'T', 'scale_A': 'None', 'scaleType_A': 'None', 'scale_B': 'None', 'scaleType_B': 'None', 'scaleBlockSize': -1, 'scaleSkipPermlane': False, 'scaleShuffleTileA': [], 'scaleShuffleTileB': []}, scaleValue_A=1, scaleValue_B=1, initMode_A='DataInitMode(Bounded)', initMode_B='DataInitMode(Bounded)', initMode_C='DataInitMode(Bounded)', workgroupMappingDim=-1, workgroupMappingValue=-1)GEMMSolution(mac_m=128, mac_n=128, mac_k=64, wave_m=32, wave_n=32, wave_k=8, wave_b=1, workgroup_size_x=128, workgroup_size_y=2, workgroupRemapXCC=False, workgroupRemapXCCValue=-1, unroll_x=0, unroll_y=0, load_A='BufferToLDSViaVGPR', load_B='BufferToLDSViaVGPR', storeLDS_D=True, betaInFma=True, scheduler='Priority', schedulerCost='LinearWeighted', prefetch=True, prefetchInFlight=2, prefetchLDSFactor=2, prefetchMixMemOps=False, loadScale_A='BufferToVGPR', loadScale_B='BufferToVGPR', swizzleScale=False, swizzleTileSize={'m': 0, 'k': 0, 'n': 0, 'l': 0}, prefetchScale=False, streamK=True, numWGs=256, streamKTwoTile=False, streamKTwoTileDPFirst=False, architecture={'ArchString': 'gfx950', 'Xnack': False, 'Sramecc': True}, matchMemoryAccess=True, version='')
- SGPR: 72 -> 75 (+3)
	| 3. FloatsGEMM(M: 7680, N: 8448, K: 8448, alpha: 2, beta: 0.5, types: {'type_A': 'half', 'type_B': 'half', 'type_C': 'half', 'type_D': 'half', 'type_acc': 'float', 'trans_A': 'N', 'trans_B': 'T', 'scale_A': 'None', 'scaleType_A': 'None', 'scale_B': 'None', 'scaleType_B': 'None', 'scaleBlockSize': -1, 'scaleSkipPermlane': False, 'scaleShuffleTileA': [], 'scaleShuffleTileB': []}, scaleValue_A: 1, scaleValue_B: 1, initMode_A: DataInitMode(Bounded), initMode_B: DataInitMode(Bounded), initMode_C: DataInitMode(Bounded), workgroupMappingDim: -1, workgroupMappingValue: -1)| GEMM(mac_m: 128, mac_n: 256, mac_k: 16, wave_m: 32, wave_n: 32, wave_k: 8, wave_b: 1, workgroup_size_x: 128, workgroup_size_y: 2, workgroupRemapXCC: False, workgroupRemapXCCValue: -1, unroll_x: 0, unroll_y: 0, load_A: BufferToLDSViaVGPR, load_B: BufferToLDSViaVGPR, storeLDS_D: True, betaInFma: True, scheduler: Priority, schedulerCost: LinearWeighted, prefetch: False, prefetchInFlight: 0, prefetchLDSFactor: 0, prefetchMixMemOps: False, loadScale_A: BufferToVGPR, loadScale_B: BufferToVGPR, swizzleScale: False, swizzleTileSize: {'m': 0, 'k': 0, 'n': 0, 'l': 0}, prefetchScale: False, streamK: True, numWGs: 256, streamKTwoTile: False, streamKTwoTileDPFirst: False, architecture: {'ArchString': 'gfx950', 'Xnack': False, 'Sramecc': True}, matchMemoryAccess: True, version: ) | GEMMProblem(M=7680, N=8448, K=8448, alpha=2, beta=0.5, types={'type_A': 'half', 'type_B': 'half', 'type_C': 'half', 'type_D': 'half', 'type_acc': 'float', 'trans_A': 'N', 'trans_B': 'T', 'scale_A': 'None', 'scaleType_A': 'None', 'scale_B': 'None', 'scaleType_B': 'None', 'scaleBlockSize': -1, 'scaleSkipPermlane': False, 'scaleShuffleTileA': [], 'scaleShuffleTileB': []}, scaleValue_A=1, scaleValue_B=1, initMode_A='DataInitMode(Bounded)', initMode_B='DataInitMode(Bounded)', initMode_C='DataInitMode(Bounded)', workgroupMappingDim=-1, workgroupMappingValue=-1)GEMMSolution(mac_m=128, mac_n=256, mac_k=16, wave_m=32, wave_n=32, wave_k=8, wave_b=1, workgroup_size_x=128, workgroup_size_y=2, workgroupRemapXCC=False, workgroupRemapXCCValue=-1, unroll_x=0, unroll_y=0, load_A='BufferToLDSViaVGPR', load_B='BufferToLDSViaVGPR', storeLDS_D=True, betaInFma=True, scheduler='Priority', schedulerCost='LinearWeighted', prefetch=False, prefetchInFlight=0, prefetchLDSFactor=0, prefetchMixMemOps=False, loadScale_A='BufferToVGPR', loadScale_B='BufferToVGPR', swizzleScale=False, swizzleTileSize={'m': 0, 'k': 0, 'n': 0, 'l': 0}, prefetchScale=False, streamK=True, numWGs=256, streamKTwoTile=False, streamKTwoTileDPFirst=False, architecture={'ArchString': 'gfx950', 'Xnack': False, 'Sramecc': True}, matchMemoryAccess=True, version='')
- SGPR: 73 -> 75 (+2) | VGPR: 154 -> 152 (-2)
	| 3. FloatsGEMM(M: 7680, N: 8448, K: 8448, alpha: 2, beta: 0.5, types: {'type_A': 'half', 'type_B': 'half', 'type_C': 'half', 'type_D': 'half', 'type_acc': 'float', 'trans_A': 'N', 'trans_B': 'T', 'scale_A': 'None', 'scaleType_A': 'None', 'scale_B': 'None', 'scaleType_B': 'None', 'scaleBlockSize': -1, 'scaleSkipPermlane': False, 'scaleShuffleTileA': [], 'scaleShuffleTileB': []}, scaleValue_A: 1, scaleValue_B: 1, initMode_A: DataInitMode(Bounded), initMode_B: DataInitMode(Bounded), initMode_C: DataInitMode(Bounded), workgroupMappingDim: -1, workgroupMappingValue: -1)| GEMM(mac_m: 128, mac_n: 256, mac_k: 16, wave_m: 32, wave_n: 32, wave_k: 8, wave_b: 1, workgroup_size_x: 128, workgroup_size_y: 2, workgroupRemapXCC: False, workgroupRemapXCCValue: -1, unroll_x: 0, unroll_y: 0, load_A: BufferToLDSViaVGPR, load_B: BufferToLDSViaVGPR, storeLDS_D: True, betaInFma: True, scheduler: Priority, schedulerCost: LinearWeighted, prefetch: False, prefetchInFlight: 0, prefetchLDSFactor: 0, prefetchMixMemOps: False, loadScale_A: BufferToVGPR, loadScale_B: BufferToVGPR, swizzleScale: False, swizzleTileSize: {'m': 0, 'k': 0, 'n': 0, 'l': 0}, prefetchScale: False, streamK: True, numWGs: 256, streamKTwoTile: False, streamKTwoTileDPFirst: True, architecture: {'ArchString': 'gfx950', 'Xnack': False, 'Sramecc': True}, matchMemoryAccess: True, version: ) | GEMMProblem(M=7680, N=8448, K=8448, alpha=2, beta=0.5, types={'type_A': 'half', 'type_B': 'half', 'type_C': 'half', 'type_D': 'half', 'type_acc': 'float', 'trans_A': 'N', 'trans_B': 'T', 'scale_A': 'None', 'scaleType_A': 'None', 'scale_B': 'None', 'scaleType_B': 'None', 'scaleBlockSize': -1, 'scaleSkipPermlane': False, 'scaleShuffleTileA': [], 'scaleShuffleTileB': []}, scaleValue_A=1, scaleValue_B=1, initMode_A='DataInitMode(Bounded)', initMode_B='DataInitMode(Bounded)', initMode_C='DataInitMode(Bounded)', workgroupMappingDim=-1, workgroupMappingValue=-1)GEMMSolution(mac_m=128, mac_n=256, mac_k=16, wave_m=32, wave_n=32, wave_k=8, wave_b=1, workgroup_size_x=128, workgroup_size_y=2, workgroupRemapXCC=False, workgroupRemapXCCValue=-1, unroll_x=0, unroll_y=0, load_A='BufferToLDSViaVGPR', load_B='BufferToLDSViaVGPR', storeLDS_D=True, betaInFma=True, scheduler='Priority', schedulerCost='LinearWeighted', prefetch=False, prefetchInFlight=0, prefetchLDSFactor=0, prefetchMixMemOps=False, loadScale_A='BufferToVGPR', loadScale_B='BufferToVGPR', swizzleScale=False, swizzleTileSize={'m': 0, 'k': 0, 'n': 0, 'l': 0}, prefetchScale=False, streamK=True, numWGs=256, streamKTwoTile=False, streamKTwoTileDPFirst=True, architecture={'ArchString': 'gfx950', 'Xnack': False, 'Sramecc': True}, matchMemoryAccess=True, version='')
- SGPR: 75 -> 76 (+1)
	| 3. FloatsGEMM(M: 7680, N: 8448, K: 8448, alpha: 2, beta: 0.5, types: {'type_A': 'half', 'type_B': 'half', 'type_C': 'half', 'type_D': 'half', 'type_acc': 'float', 'trans_A': 'N', 'trans_B': 'T', 'scale_A': 'None', 'scaleType_A': 'None', 'scale_B': 'None', 'scaleType_B': 'None', 'scaleBlockSize': -1, 'scaleSkipPermlane': False, 'scaleShuffleTileA': [], 'scaleShuffleTileB': []}, scaleValue_A: 1, scaleValue_B: 1, initMode_A: DataInitMode(Bounded), initMode_B: DataInitMode(Bounded), initMode_C: DataInitMode(Bounded), workgroupMappingDim: -1, workgroupMappingValue: -1)| GEMM(mac_m: 128, mac_n: 256, mac_k: 16, wave_m: 32, wave_n: 32, wave_k: 8, wave_b: 1, workgroup_size_x: 128, workgroup_size_y: 2, workgroupRemapXCC: False, workgroupRemapXCCValue: -1, unroll_x: 0, unroll_y: 0, load_A: BufferToLDSViaVGPR, load_B: BufferToLDSViaVGPR, storeLDS_D: True, betaInFma: True, scheduler: Priority, schedulerCost: LinearWeighted, prefetch: False, prefetchInFlight: 0, prefetchLDSFactor: 0, prefetchMixMemOps: False, loadScale_A: BufferToVGPR, loadScale_B: BufferToVGPR, swizzleScale: False, swizzleTileSize: {'m': 0, 'k': 0, 'n': 0, 'l': 0}, prefetchScale: False, streamK: True, numWGs: 256, streamKTwoTile: True, streamKTwoTileDPFirst: False, architecture: {'ArchString': 'gfx950', 'Xnack': False, 'Sramecc': True}, matchMemoryAccess: True, version: ) | GEMMProblem(M=7680, N=8448, K=8448, alpha=2, beta=0.5, types={'type_A': 'half', 'type_B': 'half', 'type_C': 'half', 'type_D': 'half', 'type_acc': 'float', 'trans_A': 'N', 'trans_B': 'T', 'scale_A': 'None', 'scaleType_A': 'None', 'scale_B': 'None', 'scaleType_B': 'None', 'scaleBlockSize': -1, 'scaleSkipPermlane': False, 'scaleShuffleTileA': [], 'scaleShuffleTileB': []}, scaleValue_A=1, scaleValue_B=1, initMode_A='DataInitMode(Bounded)', initMode_B='DataInitMode(Bounded)', initMode_C='DataInitMode(Bounded)', workgroupMappingDim=-1, workgroupMappingValue=-1)GEMMSolution(mac_m=128, mac_n=256, mac_k=16, wave_m=32, wave_n=32, wave_k=8, wave_b=1, workgroup_size_x=128, workgroup_size_y=2, workgroupRemapXCC=False, workgroupRemapXCCValue=-1, unroll_x=0, unroll_y=0, load_A='BufferToLDSViaVGPR', load_B='BufferToLDSViaVGPR', storeLDS_D=True, betaInFma=True, scheduler='Priority', schedulerCost='LinearWeighted', prefetch=False, prefetchInFlight=0, prefetchLDSFactor=0, prefetchMixMemOps=False, loadScale_A='BufferToVGPR', loadScale_B='BufferToVGPR', swizzleScale=False, swizzleTileSize={'m': 0, 'k': 0, 'n': 0, 'l': 0}, prefetchScale=False, streamK=True, numWGs=256, streamKTwoTile=True, streamKTwoTileDPFirst=False, architecture={'ArchString': 'gfx950', 'Xnack': False, 'Sramecc': True}, matchMemoryAccess=True, version='')
- SGPR: 72 -> 75 (+3)
	| 3. FloatsGEMM(M: 7680, N: 8448, K: 8448, alpha: 2, beta: 0.5, types: {'type_A': 'half', 'type_B': 'half', 'type_C': 'half', 'type_D': 'half', 'type_acc': 'float', 'trans_A': 'N', 'trans_B': 'T', 'scale_A': 'None', 'scaleType_A': 'None', 'scale_B': 'None', 'scaleType_B': 'None', 'scaleBlockSize': -1, 'scaleSkipPermlane': False, 'scaleShuffleTileA': [], 'scaleShuffleTileB': []}, scaleValue_A: 1, scaleValue_B: 1, initMode_A: DataInitMode(Bounded), initMode_B: DataInitMode(Bounded), initMode_C: DataInitMode(Bounded), workgroupMappingDim: -1, workgroupMappingValue: -1)| GEMM(mac_m: 128, mac_n: 256, mac_k: 16, wave_m: 32, wave_n: 32, wave_k: 8, wave_b: 1, workgroup_size_x: 128, workgroup_size_y: 2, workgroupRemapXCC: False, workgroupRemapXCCValue: -1, unroll_x: 0, unroll_y: 0, load_A: BufferToLDSViaVGPR, load_B: BufferToLDSViaVGPR, storeLDS_D: True, betaInFma: True, scheduler: Priority, schedulerCost: LinearWeighted, prefetch: False, prefetchInFlight: 2, prefetchLDSFactor: 0, prefetchMixMemOps: False, loadScale_A: BufferToVGPR, loadScale_B: BufferToVGPR, swizzleScale: False, swizzleTileSize: {'m': 0, 'k': 0, 'n': 0, 'l': 0}, prefetchScale: False, streamK: True, numWGs: 256, streamKTwoTile: False, streamKTwoTileDPFirst: False, architecture: {'ArchString': 'gfx950', 'Xnack': False, 'Sramecc': True}, matchMemoryAccess: True, version: ) | GEMMProblem(M=7680, N=8448, K=8448, alpha=2, beta=0.5, types={'type_A': 'half', 'type_B': 'half', 'type_C': 'half', 'type_D': 'half', 'type_acc': 'float', 'trans_A': 'N', 'trans_B': 'T', 'scale_A': 'None', 'scaleType_A': 'None', 'scale_B': 'None', 'scaleType_B': 'None', 'scaleBlockSize': -1, 'scaleSkipPermlane': False, 'scaleShuffleTileA': [], 'scaleShuffleTileB': []}, scaleValue_A=1, scaleValue_B=1, initMode_A='DataInitMode(Bounded)', initMode_B='DataInitMode(Bounded)', initMode_C='DataInitMode(Bounded)', workgroupMappingDim=-1, workgroupMappingValue=-1)GEMMSolution(mac_m=128, mac_n=256, mac_k=16, wave_m=32, wave_n=32, wave_k=8, wave_b=1, workgroup_size_x=128, workgroup_size_y=2, workgroupRemapXCC=False, workgroupRemapXCCValue=-1, unroll_x=0, unroll_y=0, load_A='BufferToLDSViaVGPR', load_B='BufferToLDSViaVGPR', storeLDS_D=True, betaInFma=True, scheduler='Priority', schedulerCost='LinearWeighted', prefetch=False, prefetchInFlight=2, prefetchLDSFactor=0, prefetchMixMemOps=False, loadScale_A='BufferToVGPR', loadScale_B='BufferToVGPR', swizzleScale=False, swizzleTileSize={'m': 0, 'k': 0, 'n': 0, 'l': 0}, prefetchScale=False, streamK=True, numWGs=256, streamKTwoTile=False, streamKTwoTileDPFirst=False, architecture={'ArchString': 'gfx950', 'Xnack': False, 'Sramecc': True}, matchMemoryAccess=True, version='')
- SGPR: 73 -> 75 (+2) | VGPR: 154 -> 152 (-2)
	| 3. FloatsGEMM(M: 7680, N: 8448, K: 8448, alpha: 2, beta: 0.5, types: {'type_A': 'half', 'type_B': 'half', 'type_C': 'half', 'type_D': 'half', 'type_acc': 'float', 'trans_A': 'N', 'trans_B': 'T', 'scale_A': 'None', 'scaleType_A': 'None', 'scale_B': 'None', 'scaleType_B': 'None', 'scaleBlockSize': -1, 'scaleSkipPermlane': False, 'scaleShuffleTileA': [], 'scaleShuffleTileB': []}, scaleValue_A: 1, scaleValue_B: 1, initMode_A: DataInitMode(Bounded), initMode_B: DataInitMode(Bounded), initMode_C: DataInitMode(Bounded), workgroupMappingDim: -1, workgroupMappingValue: -1)| GEMM(mac_m: 128, mac_n: 256, mac_k: 16, wave_m: 32, wave_n: 32, wave_k: 8, wave_b: 1, workgroup_size_x: 128, workgroup_size_y: 2, workgroupRemapXCC: False, workgroupRemapXCCValue: -1, unroll_x: 0, unroll_y: 0, load_A: BufferToLDSViaVGPR, load_B: BufferToLDSViaVGPR, storeLDS_D: True, betaInFma: True, scheduler: Priority, schedulerCost: LinearWeighted, prefetch: False, prefetchInFlight: 2, prefetchLDSFactor: 0, prefetchMixMemOps: False, loadScale_A: BufferToVGPR, loadScale_B: BufferToVGPR, swizzleScale: False, swizzleTileSize: {'m': 0, 'k': 0, 'n': 0, 'l': 0}, prefetchScale: False, streamK: True, numWGs: 256, streamKTwoTile: False, streamKTwoTileDPFirst: True, architecture: {'ArchString': 'gfx950', 'Xnack': False, 'Sramecc': True}, matchMemoryAccess: True, version: ) | GEMMProblem(M=7680, N=8448, K=8448, alpha=2, beta=0.5, types={'type_A': 'half', 'type_B': 'half', 'type_C': 'half', 'type_D': 'half', 'type_acc': 'float', 'trans_A': 'N', 'trans_B': 'T', 'scale_A': 'None', 'scaleType_A': 'None', 'scale_B': 'None', 'scaleType_B': 'None', 'scaleBlockSize': -1, 'scaleSkipPermlane': False, 'scaleShuffleTileA': [], 'scaleShuffleTileB': []}, scaleValue_A=1, scaleValue_B=1, initMode_A='DataInitMode(Bounded)', initMode_B='DataInitMode(Bounded)', initMode_C='DataInitMode(Bounded)', workgroupMappingDim=-1, workgroupMappingValue=-1)GEMMSolution(mac_m=128, mac_n=256, mac_k=16, wave_m=32, wave_n=32, wave_k=8, wave_b=1, workgroup_size_x=128, workgroup_size_y=2, workgroupRemapXCC=False, workgroupRemapXCCValue=-1, unroll_x=0, unroll_y=0, load_A='BufferToLDSViaVGPR', load_B='BufferToLDSViaVGPR', storeLDS_D=True, betaInFma=True, scheduler='Priority', schedulerCost='LinearWeighted', prefetch=False, prefetchInFlight=2, prefetchLDSFactor=0, prefetchMixMemOps=False, loadScale_A='BufferToVGPR', loadScale_B='BufferToVGPR', swizzleScale=False, swizzleTileSize={'m': 0, 'k': 0, 'n': 0, 'l': 0}, prefetchScale=False, streamK=True, numWGs=256, streamKTwoTile=False, streamKTwoTileDPFirst=True, architecture={'ArchString': 'gfx950', 'Xnack': False, 'Sramecc': True}, matchMemoryAccess=True, version='')
- SGPR: 75 -> 76 (+1)
	| 3. FloatsGEMM(M: 7680, N: 8448, K: 8448, alpha: 2, beta: 0.5, types: {'type_A': 'half', 'type_B': 'half', 'type_C': 'half', 'type_D': 'half', 'type_acc': 'float', 'trans_A': 'N', 'trans_B': 'T', 'scale_A': 'None', 'scaleType_A': 'None', 'scale_B': 'None', 'scaleType_B': 'None', 'scaleBlockSize': -1, 'scaleSkipPermlane': False, 'scaleShuffleTileA': [], 'scaleShuffleTileB': []}, scaleValue_A: 1, scaleValue_B: 1, initMode_A: DataInitMode(Bounded), initMode_B: DataInitMode(Bounded), initMode_C: DataInitMode(Bounded), workgroupMappingDim: -1, workgroupMappingValue: -1)| GEMM(mac_m: 128, mac_n: 256, mac_k: 16, wave_m: 32, wave_n: 32, wave_k: 8, wave_b: 1, workgroup_size_x: 128, workgroup_size_y: 2, workgroupRemapXCC: False, workgroupRemapXCCValue: -1, unroll_x: 0, unroll_y: 0, load_A: BufferToLDSViaVGPR, load_B: BufferToLDSViaVGPR, storeLDS_D: True, betaInFma: True, scheduler: Priority, schedulerCost: LinearWeighted, prefetch: False, prefetchInFlight: 2, prefetchLDSFactor: 0, prefetchMixMemOps: False, loadScale_A: BufferToVGPR, loadScale_B: BufferToVGPR, swizzleScale: False, swizzleTileSize: {'m': 0, 'k': 0, 'n': 0, 'l': 0}, prefetchScale: False, streamK: True, numWGs: 256, streamKTwoTile: True, streamKTwoTileDPFirst: False, architecture: {'ArchString': 'gfx950', 'Xnack': False, 'Sramecc': True}, matchMemoryAccess: True, version: ) | GEMMProblem(M=7680, N=8448, K=8448, alpha=2, beta=0.5, types={'type_A': 'half', 'type_B': 'half', 'type_C': 'half', 'type_D': 'half', 'type_acc': 'float', 'trans_A': 'N', 'trans_B': 'T', 'scale_A': 'None', 'scaleType_A': 'None', 'scale_B': 'None', 'scaleType_B': 'None', 'scaleBlockSize': -1, 'scaleSkipPermlane': False, 'scaleShuffleTileA': [], 'scaleShuffleTileB': []}, scaleValue_A=1, scaleValue_B=1, initMode_A='DataInitMode(Bounded)', initMode_B='DataInitMode(Bounded)', initMode_C='DataInitMode(Bounded)', workgroupMappingDim=-1, workgroupMappingValue=-1)GEMMSolution(mac_m=128, mac_n=256, mac_k=16, wave_m=32, wave_n=32, wave_k=8, wave_b=1, workgroup_size_x=128, workgroup_size_y=2, workgroupRemapXCC=False, workgroupRemapXCCValue=-1, unroll_x=0, unroll_y=0, load_A='BufferToLDSViaVGPR', load_B='BufferToLDSViaVGPR', storeLDS_D=True, betaInFma=True, scheduler='Priority', schedulerCost='LinearWeighted', prefetch=False, prefetchInFlight=2, prefetchLDSFactor=0, prefetchMixMemOps=False, loadScale_A='BufferToVGPR', loadScale_B='BufferToVGPR', swizzleScale=False, swizzleTileSize={'m': 0, 'k': 0, 'n': 0, 'l': 0}, prefetchScale=False, streamK=True, numWGs=256, streamKTwoTile=True, streamKTwoTileDPFirst=False, architecture={'ArchString': 'gfx950', 'Xnack': False, 'Sramecc': True}, matchMemoryAccess=True, version='')
- SGPR: 73 -> 75 (+2)

Results truncated, see full report in workspace

Links

… AddStreamK

math-ci · 2025-12-11T19:05:22Z

perfci run on commit `96262ce`

math-ci run

math-ci · 2025-12-11T23:07:57Z

perfci run on commit `7aa9059`

math-ci run

math-ci · 2025-12-12T21:00:48Z

perfci run on commit `6cb5d65`

math-ci run

yiqialiu added 9 commits December 1, 2025 23:02

Change coordinate and control grph to reset flags

3ee9b9c

Formatting

e5fc05d

Remove wait zero after reset

c096700

Map ScratchPolicy to m_scratchAllocator

5ce1959

Enable multiple scratch allocation

4a6c692

Resolve merge conflicts

d92cf90

Add unit tests

261b980

Formatting, add comments

002c05b

Minor changes

c99bb7a

github-actions Bot added project: hipblaslt shared: rocroller labels Dec 5, 2025

assistant-librarian Bot added the organization: ROCm label Dec 5, 2025

Add streamk_fp4 suite

4575181

yiqialiu added 2 commits December 6, 2025 05:56

Fix for-loop range

33e427d

Fix for-lopp range

3694e65

liu-yiqian marked this pull request as ready for review December 8, 2025 18:43

liu-yiqian requested a review from a team as a code owner December 8, 2025 18:43

liu-yiqian and others added 2 commits December 8, 2025 12:43

Merge branch 'develop' into users/liu-yiqian/reset-flag

2d7d21a

Remove streamk suite, will add that separately

fd85c81

liu-yiqian assigned ThanHenderson Dec 9, 2025

yiqialiu and others added 2 commits December 10, 2025 17:30

Update test, allocateScratch() returns size

a2979c0

Merge branch 'develop' into users/liu-yiqian/reset-flag

f8ea067

yiqialiu added 2 commits December 10, 2025 17:32

Formatting

c43dd23

Merge branch 'users/liu-yiqian/reset-flag' of github.com:ROCm/rocm-li…

ee18ee9

…braries into users/liu-yiqian/reset-flag

memmett approved these changes Dec 10, 2025

View reviewed changes

yiqialiu added 3 commits December 11, 2025 04:52

Only wave0_workitem0 writes to flag

6ea17d9

Print info when flag is not rest to 0, remove unnecessary WaitZero in…

c19d7f0

… AddStreamK

Enable rocroller streamk in hipblaslt

3b9ec64

liu-yiqian requested a review from a team as a code owner December 11, 2025 17:10

Update SolutionIndexParameters

96262ce

liu-yiqian changed the title ~~Use separate array for flags in StreamK~~ [rocRoller][hipblaslt]Use separate array for flags in rocRoller StreamK Dec 11, 2025

bnemanich reviewed Dec 11, 2025

View reviewed changes

Comment thread projects/hipblaslt/library/src/amd_detail/rocblaslt/src/rocroller/gemm.cpp Outdated

yiqialiu added 3 commits December 11, 2025 15:21

Use prob.Synchronizer for flags

fa87f64

Enable streamk

f85411b

Disable workgroupMapping when streamK is on

7aa9059

yiqialiu and others added 3 commits December 12, 2025 19:07

Not use streamk for f6 kernels

a769025

Resolve conflicts

6cb5d65

Merge branch 'develop' into users/liu-yiqian/reset-flag

90ff6bd

bnemanich approved these changes Dec 12, 2025

View reviewed changes

bnemanich merged commit d1d8188 into develop Dec 12, 2025
43 of 48 checks passed

bnemanich deleted the users/liu-yiqian/reset-flag branch December 12, 2025 21:21

jayhawk-commits mentioned this pull request Dec 16, 2025

Revert "[origami] Major refactoring of codebase (#2718)" #3411

Closed

Conversation

liu-yiqian commented Dec 5, 2025 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Motivation

Technical Details

Test Plan

Test Result

Submission Checklist

Uh oh!

math-ci Bot commented Dec 5, 2025

Generated Documentation

Uh oh!

math-ci Bot commented Dec 5, 2025

CodeQL report

Results Summary

Links

Uh oh!

math-ci Bot commented Dec 5, 2025 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Performance Report for gfx12

Results

Uh oh!

math-ci Bot commented Dec 5, 2025 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Resource Report for gfx12

Results

Uh oh!

math-ci Bot commented Dec 5, 2025 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Code Coverage Report for gfx942

Summary

Artifacts

Commit Hashes

Uh oh!

math-ci Bot commented Dec 6, 2025 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Performance Report for gfx942

Results

Uh oh!

math-ci Bot commented Dec 6, 2025 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Resource Report for gfx942

Results

Uh oh!

memmett commented Dec 9, 2025

Uh oh!

liu-yiqian commented Dec 10, 2025

Uh oh!

math-ci Bot commented Dec 10, 2025 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Performance Report for gfx950-rocroller

Results

Uh oh!

math-ci Bot commented Dec 10, 2025 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Resource Report for gfx950-rocroller

Results

Uh oh!

Uh oh!

math-ci Bot commented Dec 11, 2025

perfci run on commit 96262ce

Uh oh!

math-ci Bot commented Dec 11, 2025

perfci run on commit 7aa9059

Uh oh!

math-ci Bot commented Dec 12, 2025

perfci run on commit 6cb5d65

Uh oh!

Uh oh!

Reviewers

Assignees

Labels

Projects

Milestone

Development

Uh oh!

4 participants

liu-yiqian commented Dec 5, 2025 •

edited

Loading

math-ci Bot commented Dec 5, 2025 •

edited

Loading

math-ci Bot commented Dec 5, 2025 •

edited

Loading

math-ci Bot commented Dec 5, 2025 •

edited

Loading

math-ci Bot commented Dec 6, 2025 •

edited

Loading

math-ci Bot commented Dec 6, 2025 •

edited

Loading

math-ci Bot commented Dec 10, 2025 •

edited

Loading

math-ci Bot commented Dec 10, 2025 •

edited

Loading

perfci run on commit `96262ce`

perfci run on commit `7aa9059`

perfci run on commit `6cb5d65`