Skip to content

Pull requests: uxlfoundation/oneDNN

New pull request New
111 Open 2,650 Closed
111 Open 2,650 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
Reviews
Filter by reviews
No reviews Review required Approved review Changes requested
Assignee
Filter by who’s assigned
Assigned to nobody Loading
Sort
Sort by
Newest Oldest Most commented Least commented Recently updated Least recently updated Best match
Most reactions
👍 👎 😄 🎉 😕 ❤️ 🚀 👀

Pull requests list

gpu: conv: add early slm check to skip ngen jitting platform:gpu-intel Codeowner: @oneapi-src/onednn-gpu-intel
#4184 opened Oct 22, 2025 by ZackyLake Loading…
2 of 10 tasks
xe: sdpa: revert split Q barriers platform:gpu-intel Codeowner: @oneapi-src/onednn-gpu-intel
#4183 opened Oct 22, 2025 by syurkevi Loading…
2 tasks done
2
tests: add asynchronous Eigen threadpool runtime implementation component:build component:tests Codeowner: @oneapi-src/onednn-arch
#4182 opened Oct 21, 2025 by dzarukin Loading…
5
xe: jit: dsl: restore idx_t hash to string backport platform:gpu-intel Codeowner: @oneapi-src/onednn-gpu-intel
#4181 opened Oct 21, 2025 by kealan-barbieri Loading…
1
[rls-v3.10] xe: copy plan: adjust null dst regioning backport platform:gpu-intel Codeowner: @oneapi-src/onednn-gpu-intel
#4180 opened Oct 21, 2025 by atkassen Loading…
@atkassen
2
xe: copy plan: adjust null dst regioning platform:gpu-intel Codeowner: @oneapi-src/onednn-gpu-intel
#4179 opened Oct 21, 2025 by atkassen Loading…
@atkassen
1
x64: matmul: Enable f16 scales for AVX2 VNNI 2 platform:cpu-x64 Intel64/AMD64 processors. Codeowner: @oneapi-src/onednn-cpu-x64
#4176 opened Oct 21, 2025 by inteldimitrius Loading…
1
cpu: rv64: add rvv softmax implementation using rvv intrinsics component:common platform:cpu-rv64 RISC-V
#4173 opened Oct 21, 2025 by zhangjian29 Loading…
3 tasks done
src: gpu: intel: gemm: add integer accumulator fix to xe2 platform:gpu-intel Codeowner: @oneapi-src/onednn-gpu-intel
#4172 opened Oct 21, 2025 by hidefromkgb Loading…
1
src: gpu: intel: gemm: jit: selector: db: limited xe2 llama strats platform:gpu-intel Codeowner: @oneapi-src/onednn-gpu-intel
#4170 opened Oct 20, 2025 by hidefromkgb Loading…
2
cpu: aarch64: conv: enable jit_sve_cov for SVE 128 component:common platform:cpu-aarch64 Codeowner: @oneapi-src/onednn-cpu-aarch64
#4167 opened Oct 20, 2025 by Ryo-not-rio Loading…
2
MXFP8 example component:examples documentation A request to change/fix/improve the documentation. Codeowner: @oneapi-src/onednn-doc
#4164 opened Oct 17, 2025 by mgouicem Loading…
6
xe: concat: calculate external idx in blocked write=1 cases platform:gpu-intel Codeowner: @oneapi-src/onednn-gpu-intel
#4160 opened Oct 16, 2025 by syurkevi Loading…
2 tasks done
5
cpu: aarch64: conv: Enable SVE 128 jit uni depthwise convolutions component:common platform:cpu-aarch64 Codeowner: @oneapi-src/onednn-cpu-aarch64
#4159 opened Oct 16, 2025 by Ryo-not-rio Loading…
10
cpu: rv64: add rvv batch normalization implementation using rvv intrinsics component:common platform:cpu-rv64 RISC-V
#4158 opened Oct 16, 2025 by zhangjian29 Loading…
3 tasks done
Enable GEMV code path to brgemm matmul for avx512 using avx2 implementation platform:cpu-x64 Intel64/AMD64 processors. Codeowner: @oneapi-src/onednn-cpu-x64
#4156 opened Oct 16, 2025 by asimonov1 • Draft
[GPU] Restore conv blocking idx_t hash to string platform:gpu-intel Codeowner: @oneapi-src/onednn-gpu-intel
#4154 opened Oct 15, 2025 by kealan-barbieri Loading…
3 of 5 tasks
5
src, tests: align synchronous Eigen threadpool implementation component:build component:common component:tests Codeowner: @oneapi-src/onednn-arch
#4135 opened Oct 14, 2025 by dzarukin Loading…
7
xe: jit: gemm: use 32-bit loads for scattered quantization parameters platform:gpu-intel Codeowner: @oneapi-src/onednn-gpu-intel
#4132 opened Oct 13, 2025 by petercad Loading…
1
cpu: rv64: add rvv inner product implementation using rvv intrinsics component:common platform:cpu-rv64 RISC-V
#4115 opened Oct 11, 2025 by zhangjian29 Loading…
2 tasks done
eigen: Fix Eigen version comparison. component:tests Codeowner: @oneapi-src/onednn-arch
#4112 opened Oct 10, 2025 by cantonios Loading…
3 tasks done
3
Enable GEMV code path to brgemm matmul for f32/avx512 platform:cpu-x64 Intel64/AMD64 processors. Codeowner: @oneapi-src/onednn-cpu-x64
#4111 opened Oct 10, 2025 by asimonov1 • Draft
oneDNN v3.10 release notes backport documentation A request to change/fix/improve the documentation. Codeowner: @oneapi-src/onednn-doc
#4106 opened Oct 10, 2025 by vgvozdeva Loading…
29
cpu: aarch64: Add bf16:bf16:f32 and f32:bf16:f32 support for brdgmm convolutions component:common platform:cpu-aarch64 Codeowner: @oneapi-src/onednn-cpu-aarch64
#4099 opened Oct 9, 2025 by vishwascm Loading…
18
cpu: aarch64: remove d_type compiletime arg from JIT component:common platform:cpu-aarch64 Codeowner: @oneapi-src/onednn-cpu-aarch64
#4064 opened Oct 3, 2025 by keanucz Loading…
9
ProTip! no:milestone will show everything without a milestone.