GPUArrays.jl
GPUArrays.jl copied to clipboard
JuliaGPU
Reusable array functionality for Julia's various GPU backends.
From the definition of the struct, it seems like it should be as simple as copy-pasting it over. I can open a PR for this, but wanted to know if...
This helps flush out uses of unsupported functionality. Noticed in https://github.com/JuliaGPU/GPUArrays.jl/pull/458, where a call to `qr(::JLArray)` dispatched to `LAPACK.geqrt!` without erroring.
@lassepe I accidentally messed up your PR (#596) so I'm reopening this. I'll try to make it pass tests but if you get it fully working before then feel free...
All the breaking changes will be on the CPU backend side, the kernel interface itself is unchanged. - [ ] Replace JLBackend with POCLBackend
I find the AbstractGPUArray implementation very useful, which allows to have very generic methods for any GPU array, from [CUDA.jl](https://juliahub.com/ui/Packages/General/CUDA) to [Metal.jl](https://juliahub.com/ui/Packages/General/Metal) and others. Here, I try to implement such...
Taken from the KernelAbstractions.jl [performant matmul](https://github.com/JuliaGPU/KernelAbstractions.jl/blob/main/examples/performant_matmul.jl) example. I had to make a few changes, such as using `unsafe_indices`, since the algorithm itself does the bounds checking, and I was getting...
The following reproducer code fails with `LoadError: Scalar indexing is disallowed.`: ``` using CUDA using LinearAlgebra A = CuArray(ones(Float32, 5, 5)) # This works any(isnan, Symmetric(A)) # This does not...
Hello, I would like to use the triu! and transpose! functions on a non-contiguous view (eg. view(a', 1:2:6,4:2:8)) - is there a way make this possible (ideally for all functions...
Hello, I've recently noticed that GPUArrays don't define a `Base.hash` implementation and fallback to the default one. This requires one to `@allowscalar` which is slow and also means one has...
This may be similar to #554 MWE: ```julia using CUDA n=(16,16,16) P = CUDA.zeros(Float32,n...) active = (1:16).
