Files
composable_kernel/include
Ali Nouri 91317bdfe9 Add atomic-free MOE GEMM implementation
- Add FusedMoeGemmTilePartitioner_NoAtomic: Forces single workgroup per expert
- Add FusedMoeGemmPipelineFlatmmPolicy_NoAtomic: Fixes alignment consistency
- Update API to use no-atomic approach when intermediate_size <= Block_N0

Eliminates atomic operations by ensuring each workgroup handles complete
expert computation without K-dimension splitting.
2025-09-26 22:50:37 +00:00
..