Repository navigation
chore(squad): consolidate GEMV 6th-win + honest-ceiling campaign state - #1519
Merged
Merged
Conversation
Merge Deckard's corrected captured-replay attribution and Gaff's #1516 int8 GEMV split-K win into decisions.md; update now.md campaign brain to the 6-win / ~1.057x ORT-ahead honest-ceiling standing; append gaff/deckard histories. Remove the stale tracked inbox drop that #1516's squash landed on main (already consolidated into decisions.md). Co-authored-by: Copilot <223556219+Copilot@users.noreply.github.com>
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
1 participant
Add this suggestion to a batch that can be applied as a single commit.This suggestion is invalid because no changes were made to the code.Suggestions cannot be applied while the pull request is closed.Suggestions cannot be applied while viewing a subset of changes.Only one suggestion per line can be applied in a batch.Add this suggestion to a batch that can be applied as a single commit.Applying suggestions on deleted lines is not supported.You must change the existing code in this line in order to create a valid suggestion.Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be applied on multi-line comments.Suggestions cannot be applied while the pull request is queued to merge.Suggestion cannot be applied right now. Please check back later.
Consolidation-only state PR (no code). Persists the campaign brain + decisions log to main after PR #1516 (int8 GEMV split-K, the 6th clean token-identical win) merged.
What's in here
.squad/decisions.md— merges Deckard's corrected captured-replay attribution + Gaff's perf(cuda-ep): symmetric int8 MatMulNBits GEMV split-K on grid-starved decode projections (+9.8 tok/s) #1516 GEMV drop (22,963 → 25,954 bytes, under gate)..squad/identity/now.md— new top verdict block: 6-win arc (Fuse SSM/linear-attention reduction: drop fp32 cuDNN ReduceSum round-trip on native decode #1486/cuda_ep: fuse gated-delta linear-attn decode — route f32 L2-norm ReduceSum off cuDNN to NVRTC block reduction #1495/cuda_ep: fuse gated-delta LinearAttention gate chains (beta Sigmoid + decay Softplus) into the decode kernel (#3) #1496/perf(cuda): gate int4 f16 GEMV pipe vs plain by launch occupancy #1501/perf(cuda): warp-cooperative LinearAttention decode kernel #1503/perf(cuda-ep): symmetric int8 MatMulNBits GEMV split-K on grid-starved decode projections (+9.8 tok/s) #1516), fair gap 1.24× → ~1.057× ORT-ahead, data-shuffle retired as a dead lever, honest ceiling ~260–270 tok/s reached → kernel-chasing stopped..squad/agents/{gaff,deckard}/history.md— appended..squad/decisions/inbox/gaff-gemv-bias-fusion.md— it rode onto main via perf(cuda-ep): symmetric int8 MatMulNBits GEMV split-K on grid-starved decode projections (+9.8 tok/s) #1516's squash (tracked-inbox anti-pattern); already consolidated into decisions.md.Campaign standing
Native qwen3.5-0.8b fp16io decode: ~263.8 tok/s vs ORT eager ~279–290 = ~1.057× ORT-ahead (from 1.24× at session start). Residual is a batch=1 launch/latency floor, not a buildable kernel lever. No manufactured levers remain.
Co-authored-by: Copilot 223556219+Copilot@users.noreply.github.com