intel/graph-compiler

View on GitHub

MLIR-based toolkit targeting intel heterogeneous hardware

Stars: 54
Forks: 16
Open beginner issues: 0
Indexed issues: 51

Dominant language: C++
License: Apache-2.0
Last GitHub push: Jan 21, 2025
Latest indexed: Sep 17, 2026
Contributing guide: Contributing guide
Code of conduct: Code of conduct
Beginner labels: No beginner labels indexed
PR merge metrics: No merged PRs in 30d

Open All

51 open issues indexed

Define softmax semantics and implement a lowering to the correct sequence of named ops Open

upstream

intel/graph-compiler#10 · 21 comments · 0 reactions · 1 assignee · Indexed Sep 17, 2026
Develop a cost model based on the upstream interface Open

enhancement GPU

intel/graph-compiler#11 · 2 comments · 0 reactions · 0 assignees · Indexed Sep 17, 2026
Evaluate BRGEMM integration at LLVM IR level Open

CPU question

intel/graph-compiler#27 · 0 comments · 0 reactions · 1 assignee · Indexed Sep 17, 2026
Tensor format for input dialect Open

intel/graph-compiler#37 · 0 comments · 0 reactions · 0 assignees · Indexed Sep 17, 2026
Run tests in CI Open

ci

intel/graph-compiler#38 · 10 comments · 0 reactions · 1 assignee · Indexed Sep 17, 2026
Develop MLIR-based layout propagation passes Open

enhancement

intel/graph-compiler#42 · 1 comment · 0 reactions · 1 assignee · Indexed Sep 17, 2026
Constant folding support Open

CPU enhancement

intel/graph-compiler#56 · 0 comments · 0 reactions · 1 assignee · Indexed Sep 17, 2026
python binding and user-friendly API design Open

enhancement

intel/graph-compiler#57 · 1 comment · 0 reactions · 1 assignee · Indexed Sep 17, 2026
const weight packing support Open

CPU enhancement

intel/graph-compiler#146 · 5 comments · 0 reactions · 1 assignee · Indexed Sep 17, 2026
[Experimental] Scaled Dot Product Attention FlashAttention Algorithm Conversion Open

intel/graph-compiler#147 · 1 comment · 0 reactions · 1 assignee · Indexed Sep 17, 2026
[Experimental] Parallelize Graphs on multiple numa nodes Open

research

intel/graph-compiler#152 · 0 comments · 0 reactions · 1 assignee · Indexed Sep 17, 2026
Construct initial GPU pipeline Open

enhancement GPU

intel/graph-compiler#155 · 1 comment · 0 reactions · 1 assignee · Indexed Sep 17, 2026
Support TBB as threadpool Open

CPU enhancement

intel/graph-compiler#169 · 0 comments · 0 reactions · 0 assignees · Indexed Sep 17, 2026
Enable OV MLP benchmark in dashboard Open

ci performance

intel/graph-compiler#170 · 2 comments · 0 reactions · 0 assignees · Indexed Sep 17, 2026
Enable OV LLM benchmark in dashboard Open

ci performance

intel/graph-compiler#171 · 1 comment · 0 reactions · 0 assignees · Indexed Sep 17, 2026
Add kernel profiling support Open

CPU enhancement

intel/graph-compiler#176 · 0 comments · 0 reactions · 0 assignees · Indexed Sep 17, 2026
RoPE kernel analysis Open

performance research

intel/graph-compiler#177 · 0 comments · 0 reactions · 1 assignee · Indexed Sep 17, 2026
LLVM Translation failed for operation: builtin.unrealized_conversion_cast Open

intel/graph-compiler#218 · 0 comments · 0 reactions · 1 assignee · Indexed Sep 17, 2026
Multi-GPU support Open

intel/graph-compiler#231 · 0 comments · 0 reactions · 0 assignees · Indexed Sep 17, 2026
Evaluate OV+GC integration on CPU Open

CPU

intel/graph-compiler#238 · 6 comments · 0 reactions · 1 assignee · Indexed Sep 17, 2026
Drive oneDNN CMake modularization and integrate on demand Open

intel/graph-compiler#259 · 0 comments · 0 reactions · 0 assignees · Indexed Sep 17, 2026
XeGPU simt lowering Open

intel/graph-compiler#260 · 1 comment · 0 reactions · 1 assignee · Indexed Sep 17, 2026
Support GPU Correctness Validation in BenchGC Open

enhancement GPU validation

intel/graph-compiler#264 · 0 comments · 0 reactions · 1 assignee · Indexed Sep 17, 2026
Dynamic tensor sizes calculation in CPU part of the generated code Open

intel/graph-compiler#291 · 0 comments · 0 reactions · 0 assignees · Indexed Sep 17, 2026
Support 64-bit types for `microkernel` pipeline Open

intel/graph-compiler#292 · 0 comments · 0 reactions · 0 assignees · Indexed Sep 17, 2026
Support dynamic shape when lowering operation to vector Open

intel/graph-compiler#300 · 1 comment · 0 reactions · 1 assignee · Indexed Sep 17, 2026
Build nightly test Open

ci enhancement performance

intel/graph-compiler#317 · 3 comments · 0 reactions · 1 assignee · Indexed Sep 17, 2026
bf16 matmul's corresponding `tensor.pack` not properly optimized Open

performance

intel/graph-compiler#320 · 4 comments · 1 reaction · 1 assignee · Indexed Sep 17, 2026
Nightly Performance Result Open

intel/graph-compiler#322 · 0 comments · 0 reactions · 0 assignees · Indexed Sep 17, 2026
add tuner mode for BenchGC to support auto-tuning Open

CPU enhancement ready to review

intel/graph-compiler#330 · 1 comment · 0 reactions · 1 assignee · Indexed Sep 17, 2026
Convert large vector to physical register vector Open

CPU

intel/graph-compiler#331 · 1 comment · 0 reactions · 1 assignee · Indexed Sep 17, 2026
[`IterativeTilingAndFusionPass`] Wrap linalg.ops in a loop even if the shape is smaller than min tiling size Open

intel/graph-compiler#332 · 1 comment · 0 reactions · 0 assignees · Indexed Sep 17, 2026
Constant cache manager and runtime pipeline Open

intel/graph-compiler#341 · 0 comments · 0 reactions · 1 assignee · Indexed Sep 17, 2026
Add an option for insert gpu allocs to treat allocations as gpu-native Open

intel/graph-compiler#344 · 1 comment · 0 reactions · 1 assignee · Indexed Sep 17, 2026
Allow for passing an allocator to the GPU pipeline Open

GPU

intel/graph-compiler#345 · 0 comments · 0 reactions · 0 assignees · Indexed Sep 17, 2026
No verbose on correctness check script Open

ci ready to review validation

intel/graph-compiler#352 · 1 comment · 0 reactions · 1 assignee · Indexed Sep 17, 2026
GC check fail with clang & release build Open

bug

intel/graph-compiler#355 · 0 comments · 0 reactions · 1 assignee · Indexed Sep 17, 2026
Merge cost model Open

intel/graph-compiler#359 · 0 comments · 0 reactions · 0 assignees · Indexed Sep 17, 2026
`one-shot-bufferize` pass generates `memref.alloc()`s in GPU kernels code and breaks the pipeline Open

bug

intel/graph-compiler#360 · 5 comments · 0 reactions · 1 assignee · Indexed Sep 17, 2026
[LinalgToXeGPU] Support `linalg.matmul_transpose_a` Open

intel/graph-compiler#361 · 0 comments · 0 reactions · 0 assignees · Indexed Sep 17, 2026
Cmake-fetching GC fails Open

bug

intel/graph-compiler#371 · 0 comments · 0 reactions · 0 assignees · Indexed Sep 17, 2026
Add more test to dashboard Open

intel/graph-compiler#374 · 0 comments · 0 reactions · 1 assignee · Indexed Sep 17, 2026
`memref.alloc()` on dynamic shape tensor cannot be successfully lowered Open

intel/graph-compiler#377 · 1 comment · 0 reactions · 1 assignee · Indexed Sep 17, 2026
Single matmul runtime segfault due to K dimension out of bound access Open

bug

intel/graph-compiler#378 · 2 comments · 0 reactions · 1 assignee · Indexed Sep 17, 2026
OV single matmul case fail Open

intel/graph-compiler#382 · 1 comment · 0 reactions · 2 assignees · Indexed Sep 17, 2026
Windows build Open

intel/graph-compiler#385 · 0 comments · 0 reactions · 0 assignees · Indexed Sep 17, 2026
Python tests misconfiguration for gc-check target Open

bug

intel/graph-compiler#386 · 0 comments · 0 reactions · 0 assignees · Indexed Sep 17, 2026
Create a function test suite for OV out of OV ops translations Open

intel/graph-compiler#388 · 0 comments · 0 reactions · 1 assignee · Indexed Sep 17, 2026
Build via `compile.sh` Open

documentation

intel/graph-compiler#391 · 0 comments · 0 reactions · 0 assignees · Indexed Sep 17, 2026
`ConvertGPUXToSPIRV` fail to lower the dynamic matmul correctly Open

bug

intel/graph-compiler#403 · 0 comments · 0 reactions · 0 assignees · Indexed Sep 17, 2026

intel/graph-compiler

Get new issues in your inbox