-
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
fixup: cpu: x64: brgemm: handle src and wei scales inside the kernel
component:tests
Codeowner: @oneapi-src/onednn-arch
#3739
opened Aug 8, 2025 by
dzarukin
Loading…
[GPU] Update fwks input files
component:tests
Codeowner: @oneapi-src/onednn-arch
#3738
opened Aug 7, 2025 by
kealan-barbieri
Loading…
1 of 2 tasks
cpu: aarch64: support for destination int8 in brgemm kernel
platform:cpu-aarch64
Codeowner: @oneapi-src/onednn-cpu-aarch64
#3732
opened Aug 7, 2025 by
kasturedeeksha
Loading…
2 tasks done
cpu: aarch64: brgemm: Extend support for bf16 in BRGEMM Kernel
platform:cpu-aarch64
Codeowner: @oneapi-src/onednn-cpu-aarch64
#3731
opened Aug 7, 2025 by
vishwascm
Loading…
xe: jit: gemm: introduce protoype IR based gemm
platform:gpu-intel
Codeowner: @oneapi-src/onednn-gpu-intel
#3729
opened Aug 6, 2025 by
rjoursler
Loading…
xe: jit: use more 2D regioning in copy planner
platform:gpu-intel
Codeowner: @oneapi-src/onednn-gpu-intel
#3727
opened Aug 6, 2025 by
petercad
Loading…
scripts: synthdnn: add scatter plot report functionality
documentation
A request to change/fix/improve the documentation. Codeowner: @oneapi-src/onednn-doc
#3725
opened Aug 5, 2025 by
rjoursler
Loading…
cpu: x64: Added sprinkled prefetching
platform:cpu-x64
Intel64/AMD64 processors. Codeowner: @oneapi-src/onednn-cpu-x64
#3723
opened Aug 5, 2025 by
yair-obodovsky
Loading…
xe2: Update BMG fp16 row major M: 1-128 strategy
platform:gpu-intel
Codeowner: @oneapi-src/onednn-gpu-intel
#3719
opened Aug 4, 2025 by
CaoZhongZ
Loading…
cpu: aarch64: matmul: Enable stateless ACL lowp matmul static quant
platform:cpu-aarch64
Codeowner: @oneapi-src/onednn-cpu-aarch64
cpu: aarch64: Enable stateless ACL inner product/fully connected.
platform:cpu-aarch64
Codeowner: @oneapi-src/onednn-cpu-aarch64
[WIP] common: vtune: update primitive representation in VTune
component:common
documentation
A request to change/fix/improve the documentation. Codeowner: @oneapi-src/onednn-doc
#3714
opened Aug 4, 2025 by
avmanerikar
•
Draft
Xbyak::Reg64 extension to support more transparent save/restore operations and use APX extended GPR set
platform:cpu-x64
Intel64/AMD64 processors. Codeowner: @oneapi-src/onednn-cpu-x64
#3713
opened Aug 3, 2025 by
ankalinin
Loading…
xe: gemm: enable additional dimension reduction cases
component:common
platform:gpu-intel
Codeowner: @oneapi-src/onednn-gpu-intel
#3712
opened Aug 2, 2025 by
Simonsays095
Loading…
[WIP] Replace gemm with matmul in RNN for fp32 on avx512
platform:cpu-x64
Intel64/AMD64 processors. Codeowner: @oneapi-src/onednn-cpu-x64
cpu: conv: enable f32:f32:x8 and bf16:bf16:x8 in conv and deconv
component:common
#3705
opened Aug 1, 2025 by
asztukow
Loading…
Add OHOS judgment in the cmakelist and add OHOS macro judgment in the .h file
component:build
third_party
#3703
opened Aug 1, 2025 by
SC404
Loading…
10 tasks
tests: benchnn: graph: address some technical debts
component:graph-api
Codeowner: @oneapi-src/onednn-graph
component:tests
Codeowner: @oneapi-src/onednn-arch
#3700
opened Jul 31, 2025 by
ElaineBao
Loading…
2 tasks done
xe2: Update BMG row major M: 33-96 strategy (subdivide granularity)
platform:gpu-intel
Codeowner: @oneapi-src/onednn-gpu-intel
#3697
opened Jul 31, 2025 by
CaoZhongZ
Loading…
xe: gemm: reuse f16 strategies for bf16 problems
platform:gpu-intel
Codeowner: @oneapi-src/onednn-gpu-intel
#3693
opened Jul 30, 2025 by
Simonsays095
Loading…
common: verbose: update dispatch checks for avx512 conv implementations
component:common
platform:cpu-x64
Intel64/AMD64 processors. Codeowner: @oneapi-src/onednn-cpu-x64
#3690
opened Jul 30, 2025 by
avmanerikar
Loading…
graph: backend: dnnl: add matmul + select fusion support
component:graph-api
Codeowner: @oneapi-src/onednn-graph
component:tests
Codeowner: @oneapi-src/onednn-arch
platform:cpu-x64
Intel64/AMD64 processors. Codeowner: @oneapi-src/onednn-cpu-x64
#3683
opened Jul 29, 2025 by
xiang1guo
Loading…
cpu: x64: binary: enable per_w strategy
component:tests
Codeowner: @oneapi-src/onednn-arch
platform:cpu-x64
Intel64/AMD64 processors. Codeowner: @oneapi-src/onednn-cpu-x64
#3677
opened Jul 28, 2025 by
jstachowintel
Loading…
[WIP] f8_e4m3: saturate big values
component:common
platform:cpu-x64
Intel64/AMD64 processors. Codeowner: @oneapi-src/onednn-cpu-x64
#3675
opened Jul 27, 2025 by
ankalinin
Loading…
xe: jit: reorder: prefer word/dword for swizzle
platform:gpu-intel
Codeowner: @oneapi-src/onednn-gpu-intel
#3650
opened Jul 21, 2025 by
atkassen
Loading…
Previous Next
ProTip!
What’s not been updated in a month: updated:<2025-07-10.