Skip to content

Issues: intel/intel-xpu-backend-for-triton

New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

167 Open 1,230 Closed
167 Open 1,230 Closed
Author
Filter by author
Loading
Label
Filter by label
Loading
Use alt + click/return to exclude labels
or ⇧ + click/return for logical OR
Projects
Filter by project
Loading
Milestones
Filter by milestone
Loading
Assignee
Filter by who’s assigned
Sort
Sort by
Newest Oldest Most commented Least commented Recently updated Least recently updated Best match
Most reactions
👍 👎 😄 🎉 😕 ❤️ 🚀 👀

Issues list

Investigate how to fix ElementwiseOpConversionBase::maybeDeduplicate failing for DpasEncoding
#3577 opened Feb 28, 2025 by dchigarev
Cat+BN+Relu inference is very slow on Max 1550
#3576 opened Feb 28, 2025 by jianyizh
2
Reland upstream commit dce695e upstream: rebase PR to be up-streamed
#3573 opened Feb 28, 2025 by whitneywhtsang 2. [Triton] Upstream
1
@chengjunlu
[06-fused-attention] Determine if FP8 operand B can use 2d block load research
#3572 opened Feb 28, 2025 by whitneywhtsang
@chengjunlu
[Pytorch Upstream] python inductor/test_torchinductor_dynamic_shapes.py -k test_split_cumsum_dynamic_shapes_xpu failed upstream: pytorch
#3570 opened Feb 27, 2025 by anmyachev
1
Lunar Lake benchmarks: the profiling number not match bug Something isn't working
#3564 opened Feb 27, 2025 by kwasd
[Umbrella] Improve FlashAttention performance on default path performance umbrella
#3560 opened Feb 27, 2025 by sommerlukas 4. [Performance] Core
[Umbrella] Improve dialect infrastructure codegen: mlir umbrella
#3559 opened Feb 27, 2025 by sommerlukas 3. [Triton] Language and Runtime
[TEST] Fix test_convert_mma2mma[mma_pair0-float16-256-256] dependencies: igc tests: ut
#3556 opened Feb 27, 2025 by whitneywhtsang 3. [Triton] Language and Runtime
1
Add CUTLASS implementation to benchmarks performance research umbrella
#3555 opened Feb 26, 2025 by whitneywhtsang 4. [Performance] Core
@sommerlukas
2
[PyTorch] python test/inductor/test_aot_inductor.py AOTInductorTestABICompatibleGpu.test_triton_kernel_weird_param_order_xpu failed tests: torchinductor upstream: pytorch
#3553 opened Feb 26, 2025 by anmyachev 3. [Triton] Language and Runtime
@anmyachev
1
[PyTorch Upstream] Triton windows wheel build failed in PyTorch CICD Win server 2019 runner local builds upstream: pytorch windows
#3552 opened Feb 26, 2025 by chuanqi129 3. [Triton] Language and Runtime
@gshimansky
8
FAILED language\test_core.py::test_dot3d[8-2-64-64-64-32-32-float32-float32] - AssertionError: hw: bmg tests: ut windows
#3544 opened Feb 26, 2025 by anmyachev 3. [Triton] Language and Runtime
[RAISE-BP] Failure on tt.addptr due to unhandled arith.select codegen: mlir performance
#3543 opened Feb 26, 2025 by mfrancepillois 3. [Triton] Language and Runtime
@etiotto
1
[RAISE-BP] Failure on tt.addptr due to unhandled arith.divsi codegen: mlir performance
#3542 opened Feb 26, 2025 by mfrancepillois 3. [Triton] Language and Runtime
@etiotto
1
02-fused-softmax failed with RuntimeError: UR error error on BMG+Win hw: bmg tests: tutorials windows
#3540 opened Feb 26, 2025 by anmyachev 3. [Triton] Language and Runtime
NotADirectoryError: [WinError 267] The directory name is invalid: C:\\Users\\vagrant\\AppData\\Local\\Temp\\tmpfdi8oh5j\\results.html upstream: triton windows
#3539 opened Feb 26, 2025 by anmyachev 3. [Triton] Language and Runtime
1
@anmyachev
1
Reland upstream commit 72193bb upstream: rebase PR to be up-streamed
#3537 opened Feb 26, 2025 by whitneywhtsang 2. [Triton] Upstream
@Dewei-Wang-sh
Add FlexAttention to benchmarks/triton_kernels_benchmark codegen: attention performance
#3535 opened Feb 26, 2025 by whitneywhtsang 4. [Performance] Core
@mfrancepillois
Error when building Triton with Intel compiler on Windows bug Something isn't working local builds windows
#3530 opened Feb 25, 2025 by pbchekin 3. [Triton] Language and Runtime
1
@gshimansky
8
Fix dpas_to_block_layout_convert.mlir codegen: mlir
#3525 opened Feb 25, 2025 by whitneywhtsang 3. [Triton] Language and Runtime
PyTorch compilation failed on Windows ci local builds windows
#3522 opened Feb 25, 2025 by pbchekin 3. [Triton] Language and Runtime
[Pytorch Upstream][Regression] Pytorch CI build triton wheel failed because of out-of-memory. tests: torchinductor upstream: pytorch
#3521 opened Feb 25, 2025 by etaf 1. [PT 2.7 Upstream] TorchInductor
@alexbaden
1
[FlexAttention] get poor performance with Triton based GEMM kernel. community performance upstream: pytorch
#3518 opened Feb 25, 2025 by liangan1 5. Ecosystem enabling and performance
[dGPU] MTL, ARL, A770 failures hw: arc hw: mtl tests: tutorials tests: ut umbrella
#3511 opened Feb 24, 2025 by vlad-penkin 3. [Triton] Language and Runtime
ProTip! Updated in the last three days: updated:>2025-02-25.