jax
0a483640
- [Mosaic GPU] Make the Pallas Blackwell matmul kernel persistent.
Go
Login via GitHub
Home
Pricing
FAQ
Install
Login
via GitHub
Commit
View On
GitHub
Commit
194 days ago
[Mosaic GPU] Make the Pallas Blackwell matmul kernel persistent. There are some gains from removing the overhead from CTA scheduling. A 2-3% TC utilization improvement on a lot of configs. PiperOrigin-RevId: 776412714
References
#29312 - [Mosaic GPU] Make the Pallas Blackwell matmul kernel persistent.
#31381 - Remove old ROCm build code
#31720 - Fix ann_test.py numerical bug in target reshape
#31768 - [ROCm] Support lowering through PJRT_Triton_Extension
#32115 - Relax version requirements for ROCm Jax Plugin wheels
#33157 - Resolve undefined behavior in bitshift unit test
#33186 - Make nvidia version data optional for ROCm builds
#579 - Create rocm-test-requirements.txt
#580 - GESVDJ support for ROCm GPUs in JAX
#581 - Fix/pallas tests shared memory
#584 - Use plain bazel to test jax, use hermetic rocm dependency
#585 - update a test for checking zero ROCm GPU event
#34135 - [ROCm] update to test if there are GPU events when doing profiling on…
#587 - Mark ROCm support for gtsv2 as stable
Author
Rifur13
Committer
Google-ML-Automation
Parents
fc74a0aa
Loading