Analyze Kernel Bottleneck

Systematically identify whether a GPU kernel is compute-bound, memory-bound, or latency-bound using roofline analysis, occupancy calculations, compute/load ratio per tile, and SASS instruction inspection. Produces a decision matrix for optimization strategy selection (cp.async, warp interleaving, tiling, double-buffering, or CuAssembler hand-tuning).

pjt222 Updated 44 repo stars

File contents

pjt222/agent-almanac commit 5ec3d6bb35

Frequently asked questions

npx get-skill add pjt222/analyze-kernel-bottleneck-9