All pull requests

Pull requests

• [WIP] XeVM and XeGPU SIMT dev

  #429·akroviakov opened Jan 28, 2025Contributor·

  ·

• Bump LLVM

  #428·akroviakov opened Jan 24, 2025Contributor·

  ·
  2

• Minor fix: added missing return value assignment

  #426·AndreyPavlenko opened Jan 13, 2025Contributor·

  ·

• Add xegpu distribution patterns

  #424·kurapov-peter opened Dec 19, 2024·

  ·

• Added RoPE unit test

  #423·AndreyPavlenko opened Dec 19, 2024Contributor·

  ·

• [XeVM] Lowering to LLVM

  #416·akroviakov opened Dec 1, 2024Contributor·

  ·
  1

• Add functionality test suite for ov op translation

  #401·zhczhong opened Oct 30, 2024Contributor·

  ·
  1

  

• Add kernel time tracing support for gpu

  #381·zhczhong opened Oct 15, 2024Contributor·

  ·
  23

• OpenVino GPU single matmul test

  DO NOT MERGEready to review
  #379·WangJialei-A opened Oct 14, 2024Contributor·

  ·
  3

• [BenchGC] add tuner tools for benchgc

  ready to reviewvalidation
  #358·xurui1995 opened Sep 24, 2024Contributor·

  ·
  2

  9

  

• Avoid use DPS interface

  #356·WangJialei-A opened Sep 24, 2024Contributor·

  ·
  7

• [Transform][vector] lowering dynamic shape of tensor.pack to vector

  #351·BRUCE11111 opened Sep 20, 2024·

  ·

• [Runtime] Constant cache manager and runtime pipeline

  CPU
  #342·niuxiaog opened Sep 14, 2024Contributor·

  ·

  · 0.1 CPU - Llama functionnality support
  1

• Initial GPU pipeline

  ready to review
  #202·kurapov-peter opened Jul 30, 2024·

  ·
  1

  25

• [RFC] Constant tensors caching pass

  documentationready to review
  #183·niuxiaog opened Jul 24, 2024Contributor·

  ·

  · 0.1 CPU - General
  8

• Transform: support sdpa to flash attention kernel conversion

  #131·yifeizh2 opened Jun 13, 2024Contributor·

  ·
  1

  5

• [Transform] mlir: named op layout propagation

  ready to review
  #101·yifeizh2 opened May 27, 2024Contributor·

  ·
  1

  11

• [Transform][Vectorization] canonicalize vector with physical vector

  ready to review
  #100·BRUCE11111 opened May 27, 2024·

  ·

  · 0.1 CPU - General
  41

• [Transforms] Add constant_tensors_folding pass

  CPUready to review
  #74·niuxiaog opened May 15, 2024Contributor·

  ·

  · 0.1 CPU - Llama functionnality support
  1

  5

• [IR] Add easy-build for "if" and "for“ for general structured control flow Ops

  #62·Menooker opened May 10, 2024·

  ·
  1

• [MLIR][IR] Add easy-builder to simplify IR building in C++

  #52·Menooker opened May 9, 2024·

  ·
  4