| Name | Modified | Size | Downloads / Week |
|---|---|---|---|
| Parent folder | |||
| README.md | 2026-07-05 | 1.7 kB | |
| v2.6.0 source code.tar.gz | 2026-07-05 | 757.4 kB | |
| v2.6.0 source code.zip | 2026-07-05 | 835.3 kB | |
| Totals: 3 Items | 1.6 MB | 0 | |
AMDGPU v2.6.0
Merged pull requests:
- Setup AMD MI300 CI runners on Alps (#928) (@luraess)
- Support for WMMA instructions for RDNA4 GPUs (#929) (@ffrancesco94)
- Update Buildkite pipeline for the new JuliaGPU cluster (#930) (@maleadt)
- Add io stream to versioninfo (#932) (@luraess)
- Fix rocFFT error with multiple tasks (#935) (@luraess)
- Add precompilation workload and move SpecialFunctions to an extension (#936) (@gbaraldi)
- Update support for sparse linear algebra (#937) (@luraess)
- Add rocSPARSE ops to precompile workflow (#940) (@luraess)
- Fix segfault in qr(dA) \ b (#941) (@luraess)
- Bump actions/checkout from 6 to 7 in the all-github-actions group (#942) (@dependabot[bot])
- Add
cholesky(dA) \ bandlu(dA) \ bGPU support (#944) (@luraess) - Enable Dependabot for Julia dependencies (#949) (@vchuravy)
- Allow GPUToolbox v2 and v3 (#950) (@vchuravy)
- Update LLD_jll requirement from 15, 16, 17, 18, 19, 20 to 15, 16, 17, 18, 19, 20, 21.1 (#951) (@dependabot[bot])
- Update Statistics requirement to 1.11.1 (#953) (@dependabot[bot])
- Update LLVM_jll requirement from 15, 16, 17, 18, 19, 20 to 15, 16, 17, 18, 19, 20, 21.1 (#954) (@dependabot[bot])
- Alias WMMA_RDNA3 to WMMA (#955) (@vchuravy)
- Bump minor version (#957) (@vchuravy)
Closed issues:
- Update ROCm artifacts to newer version (#440)
- Cannot setup AMDGPU.jl (#680)
- cholesky on Hermitian matrix leads to scalar indexing error (#734)
- Scalar indexing with isapprox when we compare sparse matrices (#842)
normon ROCArrays fails because of scalar indexing (#843)- Can't allocate zeros/ones on MI300A (#938)
- Segfault in ldiv (#943)