Search results

api: introduce precomputed reductions primitive attribute

component:apiCodeowner: @oneapi-src/onednn-archCodeowner: @oneapi-src/onednn-arch

component:common

component:testsCodeowner: @oneapi-src/onednn-archCodeowner: @oneapi-src/onednn-arch

documentationA request to change/fix/improve the documentation. Codeowner: @oneapi-src/onednn-docA request to change/fix/improve the documentation. Codeowner: @oneapi-src/onednn-doc

platform:gpu-intelCodeowner: @oneapi-src/onednn-gpu-intelCodeowner: @oneapi-src/onednn-gpu-intel

Status: Open (in progress).

uxlfoundation/oneDNNnumber 3794

#3794 In uxlfoundation/oneDNN;

· dzarukin opened on Aug 20, 2025

host-side scalar scales and zps support for conv ref

component:common

component:testsCodeowner: @oneapi-src/onednn-archCodeowner: @oneapi-src/onednn-arch

platform:gpu-genericCodeowner: @oneapi-src/onednn-gpu-genericCodeowner: @oneapi-src/onednn-gpu-generic

platform:gpu-intelCodeowner: @oneapi-src/onednn-gpu-intelCodeowner: @oneapi-src/onednn-gpu-intel

Status: Draft (not ready).

uxlfoundation/oneDNNnumber 3790

#3790 In uxlfoundation/oneDNN;

· mzhukova opened on Aug 19, 2025

(new API) src: gpu: intel: add user-supplied precomputed zps to gemm

component:apiCodeowner: @oneapi-src/onednn-archCodeowner: @oneapi-src/onednn-arch

component:common

component:testsCodeowner: @oneapi-src/onednn-archCodeowner: @oneapi-src/onednn-arch

documentationA request to change/fix/improve the documentation. Codeowner: @oneapi-src/onednn-docA request to change/fix/improve the documentation. Codeowner: @oneapi-src/onednn-doc

platform:gpu-intelCodeowner: @oneapi-src/onednn-gpu-intelCodeowner: @oneapi-src/onednn-gpu-intel

Status: Open (in progress).

uxlfoundation/oneDNNnumber 3750

#3750 In uxlfoundation/oneDNN;

· hidefromkgb opened on Aug 12, 2025

[WIP] common: vtune: update primitive representation in VTune

component:common

documentationA request to change/fix/improve the documentation. Codeowner: @oneapi-src/onednn-docA request to change/fix/improve the documentation. Codeowner: @oneapi-src/onednn-doc

Status: Draft (not ready).

uxlfoundation/oneDNNnumber 3714

#3714 In uxlfoundation/oneDNN;

· avmanerikar opened on Aug 4, 2025

cpu: conv: enable f32:f32:x8 and bf16:bf16:x8 in conv and deconv

component:common

Status: Open (in progress).

uxlfoundation/oneDNNnumber 3705

#3705 In uxlfoundation/oneDNN;

· asztukow opened on Aug 1, 2025

[WIP] f8_e4m3: saturate big values

component:common

platform:cpu-x64Intel64/AMD64 processors. Codeowner: @oneapi-src/onednn-cpu-x64Intel64/AMD64 processors. Codeowner: @oneapi-src/onednn-cpu-x64

Status: Open (in progress).

uxlfoundation/oneDNNnumber 3675

#3675 In uxlfoundation/oneDNN;

· ankalinin opened on Jul 27, 2025

cpu: postops: dt support check

component:common

component:graph-apiCodeowner: @oneapi-src/onednn-graphCodeowner: @oneapi-src/onednn-graph

component:testsCodeowner: @oneapi-src/onednn-archCodeowner: @oneapi-src/onednn-arch

Status: Open (in progress).

uxlfoundation/oneDNNnumber 3634

#3634 In uxlfoundation/oneDNN;

· pmanczak opened on Jul 18, 2025

cpu: x64: pool: enable u8 type in fwd pooling

component:common

platform:cpu-x64Intel64/AMD64 processors. Codeowner: @oneapi-src/onednn-cpu-x64Intel64/AMD64 processors. Codeowner: @oneapi-src/onednn-cpu-x64

Status: Open (in progress).

uxlfoundation/oneDNNnumber 3440

#3440 In uxlfoundation/oneDNN;

· jstachowintel opened on Jun 18, 2025

src: remove unnecessary compute device info header dependencies

component:common

component:graph-apiCodeowner: @oneapi-src/onednn-graphCodeowner: @oneapi-src/onednn-graph

platform:gpu-genericCodeowner: @oneapi-src/onednn-gpu-genericCodeowner: @oneapi-src/onednn-gpu-generic

platform:gpu-intelCodeowner: @oneapi-src/onednn-gpu-intelCodeowner: @oneapi-src/onednn-gpu-intel

third_party

Status: Open (in progress).

uxlfoundation/oneDNNnumber 3428

#3428 In uxlfoundation/oneDNN;

· rjoursler opened on Jun 16, 2025

[GPU][PoC] Initial int16 support for GEMM and convolution

component:apiCodeowner: @oneapi-src/onednn-archCodeowner: @oneapi-src/onednn-arch

component:common

component:testsCodeowner: @oneapi-src/onednn-archCodeowner: @oneapi-src/onednn-arch

platform:gpu-genericCodeowner: @oneapi-src/onednn-gpu-genericCodeowner: @oneapi-src/onednn-gpu-generic

platform:gpu-intelCodeowner: @oneapi-src/onednn-gpu-intelCodeowner: @oneapi-src/onednn-gpu-intel

Status: Open (in progress).

uxlfoundation/oneDNNnumber 3419

#3419 In uxlfoundation/oneDNN;

· hidefromkgb opened on Jun 11, 2025

xe: sdpa: pass scale as a scalar kernel parameter (host side scalar memory descriptors)

component:apiCodeowner: @oneapi-src/onednn-archCodeowner: @oneapi-src/onednn-arch

component:common

component:graph-apiCodeowner: @oneapi-src/onednn-graphCodeowner: @oneapi-src/onednn-graph

component:testsCodeowner: @oneapi-src/onednn-archCodeowner: @oneapi-src/onednn-arch

platform:gpu-intelCodeowner: @oneapi-src/onednn-gpu-intelCodeowner: @oneapi-src/onednn-gpu-intel

Status: Draft (not ready).

uxlfoundation/oneDNNnumber 3412

#3412 In uxlfoundation/oneDNN;

· pv-pterab-s opened on Jun 11, 2025

cpu: aarch64: extend brgemm conv to sve_128

component:common

platform:cpu-aarch64Codeowner: @oneapi-src/onednn-cpu-aarch64Codeowner: @oneapi-src/onednn-cpu-aarch64

Status: Open (in progress).

uxlfoundation/oneDNNnumber 3363

#3363 In uxlfoundation/oneDNN;

· jondea opened on May 29, 2025