hi everyone,
I've made a few updates in the weekend and now pmmc and pmcxcl Python modules can run simulations (using OpenCL) out of box on Apple processors (M1-M4).
see my updates here
https://www.linkedin.com/feed/update/urn:li:activity:7352572814962454528/
pmcx can also be installed on Apple silicon but it only provides data processing functions such as pmcx.detweight(), pmcx.meanpath(); it can't run simulation (pmcx.run or pmcx.gpuinfo() are not available) because Apple does not support either NVIDIA GPU or CUDA.
in addition to Apple processors, pmcxcl and pmmc can also run on Intel-based Apple running OpenCL based simulations. MacOS 12 is the oldest version that they support.
Aside for MacOS support, all 3 Python modules should be installed and run properly on Linux and Windows as long as the hardware/driver is provided.
For running pmmc on Windows, as I mentioned in another thread, you need to run "pip install sparse_numba" before importing pmmc. This is not needed for either pmcx or pmcxcl.
let me know if you see any issue with these updated packages.
Qianqian