issues
search
openxla
/
xla
A machine learning compiler for GPUs, CPUs, and ML accelerators
Apache License 2.0
2.41k
stars
361
forks
source link
issues
Newest
Newest
Most commented
Recently updated
Oldest
Least commented
Least recently updated
Remove already resolved TODO.
#14405
copybara-service[bot]
closed
4 days ago
0
[xla:cpu] Pass buffer allocations for arguments and results when emitting kernel prototype
#14404
copybara-service[bot]
closed
5 days ago
0
PR #14092: NVTX: name threads, CUDA devices and CUDA streams
#14403
copybara-service[bot]
closed
4 days ago
0
PR #70765: [oneDNN] Update oneDNN library to v3.5
#14402
copybara-service[bot]
opened
5 days ago
0
[XLA:CPU] Add runtime check if `RngBitGenerator` was expanded.
#14401
copybara-service[bot]
closed
6 hours ago
1
[XLA:GPU][MLIR-based emitters] Use simplified scatters in gpu tests.
#14400
copybara-service[bot]
closed
4 days ago
0
Make NvPtxCompiler work without ptxas when libnvptxcompiler is enabled
#14399
copybara-service[bot]
closed
4 days ago
0
Move constants out of anonymous namespace
#14398
copybara-service[bot]
closed
10 hours ago
0
[XLA:GPU] Fine-grained remat policy makes async/pipelined collectives execute in the main stream
#14397
qGentry
opened
5 days ago
3
PR #14202: [fusion] Add RS->DUS dynamic slice fusion
#14396
copybara-service[bot]
opened
5 days ago
0
Vectorized multi-row reductions.
#14395
copybara-service[bot]
closed
5 days ago
0
[XLA] Fix missing build depenency.
#14394
copybara-service[bot]
closed
5 days ago
0
[XLA:GPU][MLIR-based emitters] Do not use variable names in tests directly.
#14393
copybara-service[bot]
closed
5 days ago
0
[XLA:GPU] Remove float normalization before Triton GEMM/cuBLAS rewrite
#14392
copybara-service[bot]
opened
5 days ago
0
[XLA:GPU][MLIR-based emitters] Specify atomic store alignment in bytes, not bits.
#14391
copybara-service[bot]
closed
5 days ago
0
Reverts 6e0d58792811039b66076883a59cbd716836afd6
#14390
copybara-service[bot]
closed
5 days ago
0
[XLA:GPU] Fix invalid memory dereference in `GpuCudaMallocAsyncAllocator`.
#14389
copybara-service[bot]
closed
5 days ago
0
Test for reduction indexing maps being bijections.
#14388
copybara-service[bot]
closed
5 days ago
0
Reverts 7e3f86da7241a8db61fd192b47ec0ee076fc3c0b
#14387
copybara-service[bot]
closed
5 days ago
0
[PJRT] Use AnyInvocable for WorkerThread.
#14386
copybara-service[bot]
closed
5 days ago
0
Adding translation from HLO --> StableHLO
#14385
copybara-service[bot]
closed
5 days ago
0
Integrate LLVM at llvm/llvm-project@efefee28a41e
#14384
copybara-service[bot]
closed
5 days ago
0
Extend CustomCallOp backend_config to take a DictionaryAttr
#14383
copybara-service[bot]
opened
5 days ago
0
Refactor build transforms and header replacements
#14382
copybara-service[bot]
closed
4 days ago
0
Add ulp error field for FP8 floats.
#14381
copybara-service[bot]
opened
6 days ago
0
Remove obsolete workflows from XLA and TensorFlow
#14380
copybara-service[bot]
closed
6 days ago
0
[xla:cpu] Add benchmark for compiling a chain if f32[12] buffers
#14379
copybara-service[bot]
closed
5 days ago
0
Add an XLA:CPU fusion benchmark.
#14378
copybara-service[bot]
closed
6 days ago
0
[xla:gpu] Increase the threshold for strength reducing small dots to 10M elements
#14377
copybara-service[bot]
closed
4 days ago
0
[XLA] Give NumMappedDims() internal linkage and use a simple constant instead.
#14376
copybara-service[bot]
closed
6 days ago
0
Inline buildozer rule for `stream_executor_impl`
#14375
copybara-service[bot]
closed
6 days ago
0
[XLA] Add tests to make sure that erf/erfc return results in-range
#14374
copybara-service[bot]
opened
6 days ago
0
Refactor more build target transforms
#14373
copybara-service[bot]
closed
6 days ago
0
Allow duplicate handler registration when traits and bundle addresses are equal
#14372
copybara-service[bot]
closed
5 days ago
0
[XLA] Simplify logic in hlo broadcast splitter.
#14371
copybara-service[bot]
opened
6 days ago
0
Reverts 153798d9ccad24d87a3aec76373503979a479f0c
#14370
copybara-service[bot]
closed
6 days ago
0
[XLA] Remove the special input bounder
#14369
copybara-service[bot]
closed
4 days ago
0
[XLA:CPU] Fix the detection of FP16 extension to AVX512
#14368
akhilgoe
opened
6 days ago
2
gen_gpu_hlo_compile_tests: Don't run in internal coverage infrastructure.
#14367
copybara-service[bot]
closed
5 days ago
0
Reverts 153798d9ccad24d87a3aec76373503979a479f0c
#14366
copybara-service[bot]
opened
6 days ago
0
Integrate Triton up to [a7e3476e](https://github.com/openai/triton/commits/a7e3476e86757780c2528d1100eb7aabe58d8c19)
#14365
copybara-service[bot]
closed
5 days ago
0
[ROCm] Add script to run multi gpu tests
#14364
hsharsha
opened
6 days ago
0
[XLA:CPU] Align thunks runtime to current runtime behavior for `rng` op.
#14363
copybara-service[bot]
closed
5 days ago
1
[XLA:GPU] Change `ConstraintExpression` to use `llvm::SmallVector` under the hood.
#14362
copybara-service[bot]
closed
6 days ago
0
Remove a duplicate bounds check.
#14361
copybara-service[bot]
closed
6 days ago
0
Fix file path in target patterns.
#14360
copybara-service[bot]
closed
6 days ago
0
Add Blackwell-related methods to CudaComputeCapability class
#14359
sergey-kozub
closed
5 days ago
3
Fix build failure for //xla/service/gpu/kernels:cutlass_gemm_custom_kernel_benchmarks in OSS
#14358
sergey-kozub
closed
6 days ago
0
Fix build failure for //xla/service/gpu:gpu_latency_hiding_scheduler_test in OSS
#14357
sergey-kozub
closed
6 days ago
0
Fix //xla/service/gpu:execution_stream_assignment_test in OSS
#14356
sergey-kozub
closed
6 days ago
0
Previous
Next