-
Notifications
You must be signed in to change notification settings - Fork 1.1k
Pull requests: uxlfoundation/oneDNN
Author
Label
Projects
Milestones
Reviews
Assignee
Sort
Pull requests list
xe: gemm: correct ordering of kernels
platform:gpu-intel
Codeowner: @oneapi-src/onednn-gpu-intel
#3617
opened Jul 16, 2025 by
Simonsays095
Loading…
[Xe] Introduce optimized bf16-fp8 GEMM kernels
documentation
A request to change/fix/improve the documentation. Codeowner: @oneapi-src/onednn-doc
platform:gpu-intel
Codeowner: @oneapi-src/onednn-gpu-intel
#3616
opened Jul 16, 2025 by
Simonsays095
Loading…
cpu, benchdnn: gnorm: corrected bwd compute formulae
component:common
component:tests
Codeowner: @oneapi-src/onednn-arch
platform:gpu-generic
Codeowner: @oneapi-src/onednn-gpu-generic
platform:gpu-intel
Codeowner: @oneapi-src/onednn-gpu-intel
#3615
opened Jul 16, 2025 by
dzarukin
Loading…
src: gpu: intel: sycl: do not wait on graph recording
component:common
platform:gpu-intel
Codeowner: @oneapi-src/onednn-gpu-intel
#3614
opened Jul 16, 2025 by
hidefromkgb
Loading…
[rls-v3.9] api: graph: fix scalar tensor creation and usage
backport
component:api
Codeowner: @oneapi-src/onednn-arch
component:examples
component:graph-api
Codeowner: @oneapi-src/onednn-graph
component:tests
Codeowner: @oneapi-src/onednn-arch
documentation
A request to change/fix/improve the documentation. Codeowner: @oneapi-src/onednn-doc
#3608
opened Jul 16, 2025 by
TaoLv
Loading…
[rls-v3.9] graph: backend: dnnl: enable f32 sdpa with gpu ukernel
backport
component:graph-api
Codeowner: @oneapi-src/onednn-graph
#3607
opened Jul 16, 2025 by
TaoLv
Loading…
[backport] graph: doc, interface, backend: support SDPA training
backport
component:graph-api
Codeowner: @oneapi-src/onednn-graph
component:tests
Codeowner: @oneapi-src/onednn-arch
documentation
A request to change/fix/improve the documentation. Codeowner: @oneapi-src/onednn-doc
WIP: xe: jit: refactor layout_t to support pvar_t dimensions
platform:gpu-intel
Codeowner: @oneapi-src/onednn-gpu-intel
#3604
opened Jul 16, 2025 by
echeresh
Loading…
[backport][rls-v3.9] cache: fix complex primitives creation from the blob
backport
component:common
component:graph-api
Codeowner: @oneapi-src/onednn-graph
#3603
opened Jul 16, 2025 by
dzarukin
Loading…
Fix erroneous fallthrough
component:tests
Codeowner: @oneapi-src/onednn-arch
platform:gpu-intel
Codeowner: @oneapi-src/onednn-gpu-intel
#3602
opened Jul 15, 2025 by
atkassen
Loading…
[GPU] Enable attrs with all int8 inputs
platform:gpu-intel
Codeowner: @oneapi-src/onednn-gpu-intel
#3601
opened Jul 15, 2025 by
kealan-barbieri
Loading…
3 of 4 tasks
[rls-v3.9] missing block2d definitions sdpa
backport
platform:gpu-intel
Codeowner: @oneapi-src/onednn-gpu-intel
#3600
opened Jul 15, 2025 by
syurkevi
Loading…
[rls-v3.9] xe: jit: workaround hardware read suppression bug
backport
platform:gpu-intel
Codeowner: @oneapi-src/onednn-gpu-intel
#3598
opened Jul 15, 2025 by
rjoursler
Loading…
src: gpu: intel: allow fp32 mask for xf16 SDPA in ukernel
platform:gpu-intel
Codeowner: @oneapi-src/onednn-gpu-intel
#3595
opened Jul 15, 2025 by
Jiexin-Zheng
Loading…
oneDNN v3.9 release notes
backport
documentation
A request to change/fix/improve the documentation. Codeowner: @oneapi-src/onednn-doc
#3592
opened Jul 14, 2025 by
vgvozdeva
Loading…
xe: ocl: avoid ref_gnorm fwd register spills
platform:gpu-intel
Codeowner: @oneapi-src/onednn-gpu-intel
#3588
opened Jul 11, 2025 by
rjoursler
Loading…
cpu: x64: matmul: dispatch shapes with small K to gemm
backport
platform:cpu-x64
Intel64/AMD64 processors. Codeowner: @oneapi-src/onednn-cpu-x64
#3586
opened Jul 11, 2025 by
xuxinzen
Loading…
cpu: gnorm bwd fix
component:common
component:tests
Codeowner: @oneapi-src/onednn-arch
#3584
opened Jul 11, 2025 by
raistefintel
•
Draft
2 of 4 tasks
missing block2d definitions sdpa
platform:gpu-intel
Codeowner: @oneapi-src/onednn-gpu-intel
#3581
opened Jul 11, 2025 by
syurkevi
Loading…
common: fix alignment for nested scratchpads
component:common
#3578
opened Jul 10, 2025 by
echeresh
Loading…
feat: add JIT asimd path for table-free eltwise algorithms
component:common
platform:cpu-aarch64
Codeowner: @oneapi-src/onednn-cpu-aarch64
#3572
opened Jul 10, 2025 by
focusunsink
Loading…
Strategy for QKV gemm under TP configurations
platform:gpu-intel
Codeowner: @oneapi-src/onednn-gpu-intel
#3569
opened Jul 10, 2025 by
CaoZhongZ
Loading…
[GPU] Fixup fp8 weights decomp handling
platform:gpu-intel
Codeowner: @oneapi-src/onednn-gpu-intel
#3567
opened Jul 9, 2025 by
kealan-barbieri
Loading…
2 tasks
[GPU] Fixup gemm scale dims init
platform:gpu-intel
Codeowner: @oneapi-src/onednn-gpu-intel
#3566
opened Jul 9, 2025 by
kealan-barbieri
Loading…
2 tasks done
github: workflows: Enable Coverity scan
devops
Github automation
#3559
opened Jul 9, 2025 by
vyevtyus
Loading…
Previous Next
ProTip!
What’s not been updated in a month: updated:<2025-06-17.