JuliaGPU/AMDGPU.jl: v1.2.8
Authors/Creators
- Julian Samaroo
- Anton Smirnov
- Valentin Churavy
- Alexis Montoison1
- Ludovic Räss2
- Torrance Hodgson3
- Wiktor Phillips4
- Ali Ramadhan5
- Tim Besard6
- Gabriel Baraldi6
- Julia TagBot
- Michel Schanen1
- William Bernoudy7
- Stephan Antholzer
- Juan Ignacio Polanco
- Takafumi Arakaki
- Mosè Giordano8
- Christian Guinard
- Carsten Bauer9
- jariji
- A. Leonard Nicusan10
- Utkarsh4
- Tom Hu11
- Tim Gymnich12
- Simeon David Schaub13
- Samuel Omlin14
- Oscar Smith15
- 1. Argonne National Laboratory
- 2. University of Lausanne
- 3. Curtin Institute of Data Science
- 4. @mit
- 5. @atdepth
- 6. @JuliaComputing
- 7. @dwavesystems
- 8. @UCL-ARC
- 9. German Space Agency
- 10. University of Birmingham
- 11. @codecov
- 12. AMD
- 13. KIT
- 14. CSCS - Swiss National Supercomputing Centre
- 15. JuliaHub
Description
AMDGPU v1.2.8
Bug fixes
sync_workgroupwas lacking memory barrier semantics. They have been updated to match the expectation coming from HIP.
This release has been identified as a backport. Automated changelogs for backports tend to be wildly incorrect. Therefore, the list of issues and pull requests is hidden.
<!-- **Merged pull requests:** - Regenerate wrappers for ROCm v6.3.3 (#747) (@amontoison) - Fix errors with istriu and istril (#750) (@amontoison) - Initial 1.12 enablement (#751) (@pxl-th) - More 1.12 enablement (#752) (@pxl-th) - Test doc build on CI machine (#753) (@luraess) - Fix doc build CI (#757) (@luraess) - Improve docs (#758) (@pxl-th) - Improve docs (#759) (@pxl-th) - Fixup image alignment (#760) (@luraess) - Fix fast min/max on 1.12 (#761) (@pxl-th) - Add Performance Tips to docs (#762) (@pxl-th) - Export more types (#764) (@amontoison) - HIP fix for unsafe_copy3d (#769) (@luraess) - Update indexing.jl (#771) (@amontoison) - Include AcceleratedKernels-0.4 (#772) (@anicusan) - Add code bloc to issue template (#773) (@luraess) - Improve docs (#774) (@luraess) - Add memory barrier semantics to `sync_workgroup` (#783) (@vchuravy) **Closed issues:** - unsafe_copy3d! requires 2^4 alignment (#330) - MI300X (gfx942) support for broadcast operations (#621) - multiplication of small matrices errors due to scalar indexing (#712) - AMDGPU works on normal terminal but crashes on VS Code (#749) - Compilation issues with `Flux.softmax()` and Julia v1.12.0-beta2 (#756) - AMDGPU doesn't work on ROCm 6.4 on Nobara/Fedora 42 (#763) - `collectinvokes!` error on julia 1.12.0-beta3 (#768) -->
Files
JuliaGPU/AMDGPU.jl-v1.2.8.zip
Files
(1.0 MB)
| Name | Size | Download all |
|---|---|---|
|
md5:acd538a7c3cad18676da737359c8f8db
|
1.0 MB | Preview Download |
Additional details
Related works
- Is supplement to
- Software: https://github.com/JuliaGPU/AMDGPU.jl/tree/v1.2.8 (URL)
Software
- Repository URL
- https://github.com/JuliaGPU/AMDGPU.jl