Replies: 1 comment
-
|
There were a number of reasons, but one of them was that the GPU backend was prioritized. However, we ended up building up an MLIR pipeline for that, and there are plans to migrate CPU to this new pipeline. See xla/service/gpu/fusions/mlir. |
Beta Was this translation helpful? Give feedback.
0 replies
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
-
Looking for application of MLIR in compilers I noticed a MLIR based CPU backend in older version of XLA. But it was deprecated early this year. Is there any reason why please?
Beta Was this translation helpful? Give feedback.
All reactions