xla
xla copied to clipboard
openxla
A machine learning compiler for GPUs, CPUs, and ML accelerators
[Triton] Re-enable createRemoveLayoutConversionPass
![copybara-service[bot] avatar](https://avatars.githubusercontent.com/in/44061?v=4 "copybara-service[bot] avatar"){kind=avatar}
[XLA:GPU] Support fusion of dynamic-slice into triton gemm. This change fuses dynamic-slice into triton gemm kernels provided the slice is taken along the major-most dimension, leaving all other dimensions the...
[JAX] Automatically share PGO data for GPU latency-hiding scheduler. Overall the idea is to collect profile data for each module given amount of times (which can be configured) then recompile...
It is a sub PR of https://github.com/openxla/xla/pull/9042 to add sycl platform
It is a sub PR of https://github.com/openxla/xla/pull/9042 to add spirv-llvm-translator and translation pass
Hi, I am trying to add xla customcall op for third-party hardware, but i found that the customcall process is different to tf/xla, for example, i can not find REGISTER_XLA_OP...
Currently, XLA seems supporting - CPU, GPU and TPU hardware. Does OpenXLA team plans to use Open Source OpenCL that is somewhat very similar to CUDA? Or does XLA support...
PR #10803: Expose CompiledMemoryStats on CPU and GPU Imported from GitHub PR https://github.com/openxla/xla/pull/10803 [CompiledMemoryStats](https://github.com/openxla/xla/blob/01073cf10b57c64615483545bfb34fb9ae0d3c06/xla/pjrt/pjrt_executable.h#L269-L269) is supposed to summarize the amount of memory required to run a given executable. Buffer-related fields...