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

197 Open 1,294 Closed
197 Open 1,294 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

Add FP8 GEMM example to block pointer tutorial
#3801 opened Mar 31, 2025 by alexbaden
@alexbaden
Add unit test for test_dot postop addmatrix w/ block ptr
#3799 opened Mar 31, 2025 by alexbaden
[CI] Upload all workflow attempts to the dashboards ci
#3798 opened Mar 31, 2025 by Egor-Krivov 7. CI and Tooling
@Egor-Krivov
Accuracy issue when using OPTIMIZE_REDUCTION_LOCALITY codegen: attention
#3782 opened Mar 28, 2025 by mfrancepillois 4. [Performance] Core
1
@mfrancepillois
1
Flash attention code requires more work to make it robust for use in models. codegen: attention community
#3781 opened Mar 28, 2025 by githubsgi 5. Ecosystem enabling and performance
@vlad-penkin
1
[VLLM] RuntimeError: Triton Error [ZE]: 0x78000011 happens on Ubuntu22.04 community tests: ecosystem
#3777 opened Mar 27, 2025 by yma11 5. Ecosystem enabling and performance
@vlad-penkin
3
PyTorch build failure with DLE 2025.1.0 ci dependencies: pytorch
#3769 opened Mar 26, 2025 by pbchekin 1. [PT 2.8 Upstream] TorchInductor
2
[CI][Local dev envs] Update oneAPI runtimes to 2025.1.0 ci dependencies: compiler dependencies: sycl runtime
#3768 opened Mar 26, 2025 by vlad-penkin 7. CI and Tooling
@pbchekin
1
Improve unit test coverage for DPAS layout block loads tests: ut
#3764 opened Mar 26, 2025 by alexbaden 4. [Performance] Core
1
@alexbaden
[BLOCK-PTR] Need to enhance the block io lowering for block pointer. codegen: mlir
#3762 opened Mar 26, 2025 by chengjunlu 4. [Performance] Core
@whitneywhtsang
1
[flash-linear-attention] Significant Compilation Time Discrepancy: Triton XPU (A770) vs. NVIDIA (RTX 4090) community hw: arc performance
#3761 opened Mar 26, 2025 by zhiyuan1i 5. Ecosystem enabling and performance
@alexbaden @vlad-penkin
3
check_call(...stdout=subprocess.DEVNULL)) hides the error output tests: ut
#3760 opened Mar 26, 2025 by Stonepia 3. [Triton] Language and Runtime
Change from OCL to SPV builtin codegen: dpas codegen: gemm dependencies: igc performance
#3754 opened Mar 25, 2025 by whitneywhtsang 4. [Performance] Core
@whitneywhtsang
Change 2D block prefetch from OCL to SPV builtin dependencies: igc
#3753 opened Mar 25, 2025 by whitneywhtsang
1
@whitneywhtsang
Change 2D block store from OCL to SPV builtin dependencies: igc
#3752 opened Mar 25, 2025 by whitneywhtsang
1
@whitneywhtsang
Support 32 threads-per-warp for DPAS layout encoding codegen: mlir performance
#3750 opened Mar 25, 2025 by LiyangLingIntel 4. [Performance] Core
[SGLANG] test_context_attention gets seg fault when is_causal=True codegen: attention tests: ecosystem
#3749 opened Mar 25, 2025 by LiyangLingIntel 5. Ecosystem enabling and performance
@LiyangLingIntel
1
[SGLANG] test_extend_attention does not produce correct results codegen: attention tests: ecosystem
#3748 opened Mar 25, 2025 by LiyangLingIntel 5. Ecosystem enabling and performance
@LiyangLingIntel
Removal of block pointer dependency codegen: mlir
#3746 opened Mar 24, 2025 by etiotto 5. Ecosystem enabling and performance
1
[ktransformers] Initial enabling tests: ecosystem
#3743 opened Mar 24, 2025 by vlad-penkin 5. Ecosystem enabling and performance
[FA Performance] Add support for split barriers codegen: attention
#3742 opened Mar 24, 2025 by mfrancepillois 4. [Performance] Core
1
@mfrancepillois
Change 2D block load from OCL to SPV builtin codegen: gemm dependencies: igc performance
#3741 opened Mar 24, 2025 by whitneywhtsang 4. [Performance] Core
1
@whitneywhtsang
[FA Performance] Optimize subgroup shape when num-warps>8 codegen: attention performance
#3740 opened Mar 24, 2025 by mfrancepillois 4. [Performance] Core
@mfrancepillois
1
[ktransformers] The attached triton code crashed on Intel PVC community tests: ecosystem
#3739 opened Mar 24, 2025 by w-chen-cc 5. Ecosystem enabling and performance
@vlad-penkin
Change DPAS from OCL to SPV builtin dependencies: igc
#3731 opened Mar 21, 2025 by whitneywhtsang 3. [Triton] Language and Runtime
1
@whitneywhtsang
1
ProTip! Add no:assignee to see everything that’s not assigned.