Search results

[TEST] Fix language/test_core.py::test_abs_fp8

Status: Open.

#6326 In intel/intel-xpu-backend-for-triton;

· whitneywhtsang opened on Mar 10, 2026

Investigate torch-free runtime dispatch for XPU backend (inspired by upstream triton-lang/triton#9578)

research

Status: Open.

#6324 In intel/intel-xpu-backend-for-triton;

· wdziurdz opened on Mar 10, 2026

Test is not skipped but mentioned in skiplist

ci

tests: ut

Status: Open.

Bug

#6323 In intel/intel-xpu-backend-for-triton;

· kwasd opened on Mar 10, 2026

[VLLM] Update VLLM pin

Status: Open.

Task

#6318 In intel/intel-xpu-backend-for-triton;

· Egor-Krivov opened on Mar 10, 2026

·  

5. Ecosystem enabling and performance

w8a8 block quant fp8 matmul in triton on B60/B580 gives incorrect result compared to CUDA

Status: Open.

#6317 In intel/intel-xpu-backend-for-triton;

· polisettyvarma opened on Mar 10, 2026

[SYCL] Use work group scratch memory extension for kernels requiring shared memory

performance

triton: runtime

Status: Open.

Task

#6315 In intel/intel-xpu-backend-for-triton;

· uditagarwal97 opened on Mar 10, 2026

·  

3. [Triton] Language and Runtime

[triton-ext] Enable Triton extensions on Windows

research

windows

Status: Open.

Feature

#6291 In intel/intel-xpu-backend-for-triton;

· quinnlp opened on Mar 6, 2026

·  

3. [Triton] Language and Runtime

[pytorch upstream] sqrt_rn accuracy may have issue when it's fused with other ops

accuracy

dependencies: igc

tests: torchinductor

upstream: pytorch

Status: Open.

Bug

#6261 In intel/intel-xpu-backend-for-triton;

· jianyizh opened on Mar 4, 2026

·  

1. [PT 2.11] TorchInductor

Inductor tests: Failed to parse and upload json test reports: Unable to locate credentials

ci

tests: torchinductor

Status: Open.

Bug

#6254 In intel/intel-xpu-backend-for-triton;

· kwasd opened on Mar 3, 2026

·  

7. CI and Tooling

A770 (Windows only): Not all tests to skip exist: test_06_fused_attention[fp8_only,skip_fp8]

hw: arc

tests: tutorials

Status: Open.

Bug

#6249 In intel/intel-xpu-backend-for-triton;

· kwasd opened on Mar 3, 2026

·  

7. CI and Tooling

[FlexAttention] Performance optimization for FlexAttention op benchmark.

codegen: attention

dependencies: pytorch

performance

Status: Open.

Task

#6245 In intel/intel-xpu-backend-for-triton;

· liangan1 opened on Mar 3, 2026

·  

4. [Performance] Core

[FlexAttention] Performance optmization for LLM inference

codegen: attention

performance

upstream: pytorch

Status: Open.

Task

#6244 In intel/intel-xpu-backend-for-triton;

· liangan1 opened on Mar 3, 2026

·  

4. [Performance] Core