Arun Thangamani 8390909842

[mlir][x86vector] Lower BF16 vector.contract to FMA using AVX2 BF16 packed ops. (#170267) ...

A `transform` pass to lower `BF16` type `vector.contract` to
`vector.fma` using `AVX2` BF16 packed operations:

- `vbcstnebf162ps` - Broadcasts BF16 into packed F32.
- `vcvtneebf162ps` - Convert packed BF16 even-indexed elements into
packed F32.
- `vcvtneobf162ps` - Convert packed BF16 odd-indexed elements into
packed F32 Data.

2025-12-17 14:41:58 +01:00

..

CMakeLists.txt

…

X86VectorUtils.cpp

…