tree c51889cbbdbf227857f890dc7854b618de952660
parent f71dd1269d3ec54c07e5d77eada3f4a9ab7cb66f
author Stanley Winata <68087699+raikonenfnu@users.noreply.github.com> 1730767106 -0800
committer GitHub <noreply@github.com> 1730767106 -0800
gpgsig -----BEGIN PGP SIGNATURE-----
 
 wsFcBAABCAAQBQJnKWkCCRC1aQ7uu5UhlAAAkWIQABuOhASo03g9D77REL8hVI0Y
 4KCX2C/TS6+LsBpjnXDrqPFPHwiiZ4gcGZaOPeDpNJ4kp8JKRPKcgEBAEcrxg35E
 ejvBWgTCArLbKaPb+xFsi9sjy8L1Tnsy0hK/t2J2ZeUjxPzlQbuGhmCZFkP5TpRH
 wuSvbxHDR4QKGYNk74r67F4plgr+9aWdKSxbHwKz4aQAiEDANRfkNMMhfS5gPDDi
 iz0fYLS3nSaBLmO1+u5GqhDzDS7UN6k0wJVcOv1nHKo6C82sLJ7fnDFUCrwFHALj
 6u3/qhQWBL0Ba/kxjlGTWLWIkuxyVzWJb1konhKMa92Y4iah4YPbY9l8QF4i7n3p
 Eb7MUTFJh41vBeT3sx5KJIjAjC6pLuLZF4zINpL5zq2d6eSqDMLVpsOupfDk8Rwj
 8M4YBq37M1SZl9lOBgYOuMN7WRxlR9Wtl+3OeyG94i6lBll6hCJYiCOqpTjjJM9J
 /sS5QE8G5aE5Z7XV31Ix14nAnkJyEk2K+rxB2be6F/w82XESCL/398vRKMzy4t7R
 Ga7eyjXj3lVVpiFeA2J9NtUIbruNUi1rIl6KN/iVuInsocJN80egC9fwnET6ac08
 yq2DaUtDO0tavUz4LeNSUxgkQKlSoSdQGgKYwaT2acPGncVeSw76SPQj+NaAScRg
 sUA3Kn2HTxcTdTNqvqhC
 =io7B
 -----END PGP SIGNATURE-----
 

[LLVMGPU] Add VMFMA for FP8 to align layouts between chained F8 contractions. (#19020)

This PR introduces virtual intrinsics on F8 MFMA that breaks apart a
single 8xF8 read into two interleaved 4xF8 read from shared memory.

This main motivation for this virtual intrinsic is to enable faster F8
attention/chained matmuls. The reason for that is by doing interleaved
reads on K-dimension, we can match the native F8 intrisic output layout
coming from the 1st matmul to the rhs read of the 2nd matmul(with
interleaved virtual MFMA layout).

Once the layout is aligned, we just need to handle it using to_layout
lowering that does reshape on the SIMT vector.

This PR has been tested on attention of shape:
[B: 1, M: 4096, K1: 64, K2: 4096, N: 64]

as seen in this IR:
[(link)](https://gist.githubusercontent.com/raikonenfnu/4d33b5addfa9c4ec9e76918704251e39/raw/5b20c0c359e3e2df7f8db4890d3cc0590352d18a/attention_f8_perf.mlir)
 
and using this spec to specify the VMFMA on 2nd matmul and regular MFMA
on 1st matmul:
([link](https://gist.githubusercontent.com/raikonenfnu/4d33b5addfa9c4ec9e76918704251e39/raw/5b20c0c359e3e2df7f8db4890d3cc0590352d18a/attn_config.mlir))

we were able to get perf of 1.63x speed up from (reference with same
config but using MFMA_16x16x32xF16 on both contractions. With
correct/same numerics.

Signed-off-by: Stanley Winata <stanley.winata@amd.com>