GPU not being used with PyOpenCL/PyCUDA even when explicitly running it

[Loci Skinner]

Hi all,

I'm quite new to PySPH and have been attempting to get it to run either the PYSPH examples or my own PySPH script using PyOpenCL and PyCUDA (as I think both were required?). I believe everything has installed and/or built correctly and when adding "--opencl" to the command line I get the extra choice to pick which platform to run on and it appears to go through all the set up correctly.

However, the issue is that it simply doesn't make use of my GPU (RTX 4070 Super)? There is a short blip of GPU usage I can see on setup, and for about half a second after setup, as the sim starts to run, but after that my GPU sits entirely unused. I get a couple warnings about implicitly defined variables right before it runs, but it doesn't error out or anything.

I've confirmed OpenCL is working correctly (I think) using this benchmark: https://github.com/ProjectPhysX/OpenCL-Benchmark .

Am I missing something here? Any help at all is appreciated, let me know if there's anything else you might need to know and I'll try provide it.

Thanks,

Loci