Merge commit '1b1c46e508c1fd40a03f54114b6b78629032fb4f' into develop

This commit is contained in:
assistant-librarian[bot]
2025-11-11 17:12:49 +00:00
parent 0b000816a4
commit db12c41b56
65 changed files with 845 additions and 455 deletions

View File

@@ -240,7 +240,10 @@ struct QuantGemmKernel
return dim3(TilePartitioner::GridSize(M, N), 1, KBatch);
}
CK_TILE_HOST static constexpr auto BlockSize() { return dim3(kBlockSize); }
CK_TILE_HOST static auto BlockSize()
{
return is_wave32() ? dim3(kBlockSize / 2) : dim3(kBlockSize);
}
CK_TILE_HOST static constexpr QuantGemmKernelArgs
MakeKernelArgs(const QuantGemmHostArgs& hostArgs)