I’m pretty new to writing code for the GPU. I’m using the variable MTLComputePipelineState.maxTotalThreadsPerThreadgroup to get the number of threads I can use. On my Intel MacBook with an intel chip and internal graphics I get 1024. On my other machine (with apple M1) I also get 1024. Doesn't this variable tell me how many threads the machine can run in parallel so shouldn’t the number be higher on a newer machine?
I tried to find information online but I couldn’t.