Hello, is GPU parallelisation reserved to CUDarray/NVIDIA cards, or is there an API for generic (i.e. Intel integrated) cards ?
And if yes, is it worth it? Where could I get more doc/info about it ?
You can use OpenCL.jl to execute (hand written) OpenCL kernels on Intel CPUs and GPUs. However, the high level CLArrays are not working on Julia v1.n at the moment.
If you use an integrated GPU, you will not necessarily get a speedup compared to the CPU (running code in parallel).
Once Intel upstreams their LLVM GPU changes, we will be able to improve this situation more easily by taking the same approach as CUDAnative does (using the appropriate LLVM backend).