Why it matters
FA-2 is standard attention. Understanding shapes performance.
Advertisement
The architecture
Tiling: work on blocks in SRAM.
Reduce non-matmul FLOPs.
Better sequence-parallel.
Advertisement
How it works end to end
Reduce operations that don't use tensor cores.
Better warp partitioning.
Sequence-parallel outer loop.
Result: 2x speedup, 70% peak FLOPs on A100.