tree e4d3c786e1a60ee490cf2be66e96fe734ca96d58
parent 0d81062a8ffabf3561e9d6cc758b8250b2f607d3
author Lei Zhang <antiagainst@google.com> 1685653710 -0400
committer GitHub <noreply@github.com> 1685653710 +0000
gpgsig -----BEGIN PGP SIGNATURE-----
 
 wsBcBAABCAAQBQJkeQjOCRBK7hj4Ov3rIwAAiEMIAKwwTwHf8egs0ZpFldw6gAi9
 FGkF/5GaSxE6JWUgKmbIr2EchISkJPaI07ncpDj6+OKeU2PB1PX4ngwPwA/hVRKh
 AKW7AQhVENz+g0yWnuvCXBQmV9UX6rXYrWDvkpCkmfxC27aqSyM06ajCFxhWmest
 Eycl44NkwkXLaNoaDnF2u6438AMGytORWxKWczkQxU+PdOygbAXdvlE+HbrqXV6R
 OLyZx+BL0gP635pXjShixsuCDzciM1q0esF4FjUdAdgsQExIeWDUkFufckv5031V
 JRUhOhMvoUNst/GhS1kmkyaAvzxWoSgRA5DAaYyISRJGL0JupdEo9qPLXcIr6b8=
 =hdK9
 -----END PGP SIGNATURE-----
 

[gpu] Enable fusing input producers after tiling reduction loops (#13806)

This commit enables fusing input producers into the
serial loop after tiling the matmul K dimension. 
This enables using the GPU shared memory promoted
for A/B matrix slices for weight dequantization, without
introducing additional shared memory allocations
during bufferization. This is to support int4 weight
quantized matmuls in LLMs.