-
Notifications
You must be signed in to change notification settings - Fork 987
Pull requests: oneapi-src/oneDNN
Author
Label
Projects
Milestones
Reviews
Assignee
Sort
Pull requests list
generic: sycl: binary: bugfix post ops with blocked formats
platform:gpu-generic
Codeowner: @oneapi-src/onednn-gpu-generic
#2105
opened Sep 20, 2024 by
t4c1
Loading…
Re-enablement of bf16 jit:uni reorders for aarch64
#2104
opened Sep 20, 2024 by
Shreyas-fuj
Loading…
Changed blocking heuristics for M,K,N in BRGEMM Matmul to enable better memory/thread utilization in aarch64
#2103
opened Sep 20, 2024 by
Shreyas-fuj
Loading…
common: replace loop with algorithm for primitive_hashing.cpp
#2102
opened Sep 19, 2024 by
Permanence-AI-Coder
Loading…
2 tasks done
rfc: dnncompat compatibility layer
RFC
A design document
#2101
opened Sep 19, 2024 by
sgeor255
Loading…
cpu: aarch64: Expand brgemm aarch64 unsupported cases handling mechanism
platform:cpu-aarch64
Codeowner: @oneapi-src/onednn-cpu-aarch64
#2099
opened Sep 18, 2024 by
Radu2k
Loading…
4 tasks done
src: cpu: aarch64: injectors: eltwise: Improve performance of tanh for block size 16
platform:cpu-aarch64
Codeowner: @oneapi-src/onednn-cpu-aarch64
cpu: x64: brgemm: use range-based loop for jit_brgemm_amx_uker.cpp
platform:cpu-x64
Intel64/AMD64 processors. Codeowner: @oneapi-src/onednn-cpu-x64
cpu: {x64,aarch64}: fix debug verbose output
platform:cpu-aarch64
Codeowner: @oneapi-src/onednn-cpu-aarch64
platform:cpu-x64
Intel64/AMD64 processors. Codeowner: @oneapi-src/onednn-cpu-x64
cpu: x64: brgemm: use range-based loop
platform:cpu-x64
Intel64/AMD64 processors. Codeowner: @oneapi-src/onednn-cpu-x64
gpu: generic: deconvolution: move implementation from ocl folder
platform:gpu-generic
Codeowner: @oneapi-src/onednn-gpu-generic
platform:gpu-intel
Codeowner: @oneapi-src/onednn-gpu-intel
src: cpu: aarch64: injectors: eltwise_injector - improve gelu performance for block size 16
platform:cpu-aarch64
Codeowner: @oneapi-src/onednn-cpu-aarch64
cpu: aarch64: allow sbgemm config for matmul primitive
platform:cpu-aarch64
Codeowner: @oneapi-src/onednn-cpu-aarch64
Start enabling the "generic" OpenCL vendor
platform:gpu-generic
Codeowner: @oneapi-src/onednn-gpu-generic
Fixed parser of OpenCL 3.0 for NVIDIA GPU
platform:gpu-nvidia
Codeowner: @oneapi-src/onednn-gpu-nvidia
#2001
opened Jul 20, 2024 by
redradist
Loading…
rfcs: add proposal for collecting and analyzing synthetic data
RFC
A design document
#1999
opened Jul 18, 2024 by
rjoursler
Loading…
gpu: intel: Optimize reusable layer normalization using work-group based reductions
performance
platform:gpu-intel
Codeowner: @oneapi-src/onednn-gpu-intel
#1990
opened Jul 10, 2024 by
umar456
Loading…
cpu: add loongarch64 support
platform:cpu-loongarch64
#1923
opened May 20, 2024 by
loong-hy
Loading…
[DO NOT REVIEW][TEST] common: on-purpose check removal to trigger new coverity issues
#1896
opened May 7, 2024 by
dzarukin
Loading…
Previous Next
ProTip!
Updated in the last three days: updated:>2024-09-19.