Skip to content

Use max clock for analytical calculations of peak flops #2801

New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Merged
jlebar merged 1 commit into from
Dec 14, 2023
Merged

Use max clock for analytical calculations of peak flops #2801

Changes from all commits
Commits
Show all changes
1 commit
Select commit
File filter

Filter by extension

Filter by extension
Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
4 changes: 2 additions & 2 deletions python/triton/ops/matmul_perf_model.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -12,7 +12,7 @@ def get_tensorcore_tflops(backend, device, num_ctas, num_warps, dtype):
''' return compute throughput in TOPS '''
total_warps = num_ctas * min(num_warps, 4)
num_subcores = driver.utils.get_device_properties(device)["multiprocessor_count"] * 4 # on recent GPUs
cur_sm_clock = nvsmi(['clocks.current.sm'])[0]
cur_sm_clock = nvsmi(['clocks.max.sm'])[0]
tflops = min(num_subcores, total_warps) / num_subcores * get_max_tensorcore_tflops(
dtype, cur_sm_clock, backend, device)
return tflops
Expand All @@ -22,7 +22,7 @@ def get_simd_tflops(backend, device, num_ctas, num_warps, dtype):
''' return compute throughput in TOPS '''
total_warps = num_ctas * min(num_warps, 4)
num_subcores = driver.utils.get_device_properties(device)["multiprocessor_count"] * 4 # on recent GPUs
cur_sm_clock = nvsmi(['clocks.current.sm'])[0]
cur_sm_clock = nvsmi(['clocks.max.sm'])[0]
tflops = min(num_subcores, total_warps) / num_subcores * get_max_simd_tflops(dtype, cur_sm_clock, backend, device)
return tflops

Expand Down