Skip to content

Pull requests: uxlfoundation/oneDNN

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
Assignee
Filter by who’s assigned
Assigned to nobody Loading
Sort

Pull requests list

cpu: x64: enable hf8 support for conv, deconv on AVX10.2 component:common platform:cpu-x64 Intel64/AMD64 processors. Codeowner: @oneapi-src/onednn-cpu-x64
#3327 opened May 23, 2025 by xuxinzen Loading…
various improvements for examples documentation component:examples documentation A request to change/fix/improve the documentation. Codeowner: @oneapi-src/onednn-doc
#3326 opened May 23, 2025 by mzhukova Draft
gpu: nvidia: amd: Remove mode modifier P for native backends component:tests Codeowner: @oneapi-src/onednn-arch
#3325 opened May 23, 2025 by ShanoToni Loading…
gpu: nvidia: Add check for deconv bias mixed dt platform:gpu-nvidia Codeowner: @oneapi-src/onednn-gpu-nvidia
#3324 opened May 23, 2025 by ShanoToni Loading…
xe: sdpa: add support for reusable sdpa [WIP] component:tests Codeowner: @oneapi-src/onednn-arch platform:gpu-intel Codeowner: @oneapi-src/onednn-gpu-intel
#3322 opened May 23, 2025 by syurkevi Loading…
2 tasks
synthdnn: add run collection type
#3320 opened May 22, 2025 by skazakov1 Loading…
xe: softmax: restore missing inf_as_zero functionality platform:gpu-intel Codeowner: @oneapi-src/onednn-gpu-intel
#3318 opened May 22, 2025 by pv-pterab-s Draft
synthdnn: generalize type interface
#3317 opened May 22, 2025 by Simonsays095 Loading…
xe: jit: gemm: group binary postop loads platform:gpu-intel Codeowner: @oneapi-src/onednn-gpu-intel
#3316 opened May 22, 2025 by Simonsays095 Loading…
generic: sycl: Extend indexing fix to BWD convolutions platform:gpu-generic Codeowner: @oneapi-src/onednn-gpu-generic
#3314 opened May 22, 2025 by ShanoToni Loading…
Revert "cpu: aarch64: Nuke winograd" component:common component:tests Codeowner: @oneapi-src/onednn-arch documentation A request to change/fix/improve the documentation. Codeowner: @oneapi-src/onednn-doc platform:cpu-aarch64 Codeowner: @oneapi-src/onednn-cpu-aarch64
#3310 opened May 21, 2025 by Ryo-not-rio Draft
perf: add ctime comparison to benchdnn_comparison.py devops Github automation
#3309 opened May 21, 2025 by gausah-arm Loading…
Extend JIT Depthwise Convolution to BF16 component:common platform:cpu-aarch64 Codeowner: @oneapi-src/onednn-cpu-aarch64
#3308 opened May 21, 2025 by renato-arantes Loading…
3 tasks done
Pmanczak/matmul goes into ref path with mask 12 component:common component:tests Codeowner: @oneapi-src/onednn-arch platform:cpu-x64 Intel64/AMD64 processors. Codeowner: @oneapi-src/onednn-cpu-x64
#3307 opened May 21, 2025 by pmanczak Loading…
rfcs: asynchronous CPU runtime support RFC A design document
#3305 opened May 20, 2025 by mgouicem Loading… v3.9
ci: aarch64: update regression tests devops Github automation
#3299 opened May 20, 2025 by Ryo-not-rio Loading…
graph: backend: dnnl: clean up macros and env vars component:graph-api Codeowner: @oneapi-src/onednn-graph
#3297 opened May 20, 2025 by TaoLv Loading…
Update dynamic quant int4 support documentation A request to change/fix/improve the documentation. Codeowner: @oneapi-src/onednn-doc
#3295 opened May 19, 2025 by kealan-barbieri Loading…
2 tasks
xe: jit: gemm: post_ops restrict mad simd size on XeHPG platform:gpu-intel Codeowner: @oneapi-src/onednn-gpu-intel
#3294 opened May 19, 2025 by dyoussif Loading…
cpu: x64: reorder: change couting mask product platform:cpu-x64 Intel64/AMD64 processors. Codeowner: @oneapi-src/onednn-cpu-x64
#3292 opened May 19, 2025 by jstachowintel Draft
Fix performance regressions in RNN platform:cpu-x64 Intel64/AMD64 processors. Codeowner: @oneapi-src/onednn-cpu-x64
#3291 opened May 19, 2025 by asimonov1 Draft
[POC] xe: jit: gemm: nf4 weights decompression component:api Codeowner: @oneapi-src/onednn-arch component:common component:tests Codeowner: @oneapi-src/onednn-arch platform:gpu-intel Codeowner: @oneapi-src/onednn-gpu-intel
#3287 opened May 17, 2025 by petercad Draft
graph: utils: pm: correctly handle empty optional subgraph and multi-consumer one component:graph-api Codeowner: @oneapi-src/onednn-graph component:tests Codeowner: @oneapi-src/onednn-arch
#3282 opened May 16, 2025 by ShanSimu Loading…
[GPU] Fixup GEMM strategy override platform:gpu-intel Codeowner: @oneapi-src/onednn-gpu-intel
#3281 opened May 15, 2025 by kealan-barbieri Loading…
1 of 2 tasks
ProTip! What’s not been updated in a month: updated:<2025-04-23.