I didn't add the synchronize function since I saw that any implementation doesn't have one. I'm also wondering why since the output might be incorrect until we wait the kernel.

Maybe this can also fix #558 ?

I didn't add the synchronize function since I saw that any implementation doesn't have one. I'm also wondering why since the output might be incorrect until we wait the kernel.

Because with Julia all kernels execute asynchronously (from the CPU perspective) and in a stream-ordered fashion (on the GPU).

Synchronization happens either explicitly by user calling CUDA.synchronize() or implicitly during GPU->CPU transfer Array(gpu_array).

So adding explicit synchronization is redundant and would only slow down GPU.

But what if an external user does for example c = kron(a, b); d = c * a?

If it is asynchronous, c may be not updated and d would be uncorrected. Am I wrong?

The user may explicit call synchronize, but the code is in principle type-agnostic, such that it could work on CPU or in the GPU as well.

But what if an external user does for example c = kron(a, b); d = c * a?

If it is asynchronous, c may be not updated and d would be uncorrected. Am I wrong?

The user may explicit call synchronize, but the code is in principle type-agnostic, such that it could work on CPU or in the GPU as well.

GPU executes kernels in order of their submission. So GPU will first compute kron and only when it is done c * a.
That is true when kernels are executed on the same stream (which is the default).

Ok very clear thanks

Here I'm indexing on C instead of A or B. This should be easier and faster, I did some benchmarks.

The CUDA implementation has the stride stuff, which I never understood if I should use it also for KernelAbstractions.

The CUDA implementation has the stride stuff, which I never understood if I should use it also for KernelAbstractions.

Didn't you implement that kernel in CUDA.jl? JuliaGPU/CUDA.jl#2177

Yes. I mean that this implementation should be faster and simpler.

Assuming that we implement the same also for the CUDA case, I still don't understand if I need to use the stride stuff also for KernelAbstractions. From the examples I have seen so far, it seems that it is handled internally in KernelAbstractions.