For those interested in contributing, there are a number of things we wanted to look into if we had the time:

  1. Testing the kmeans pytorch CPU vs GPU test code on a fresh Manjaro install with Nvidia proprietary drivers to verify consistent results.
  2. Test other algorithims. We only used the kmeans pytorch example code. There are lots of other machine learning tests and also lots of CUDA acceleratable workloads outside of machine learning.
  3. Testing OpenCL and OpenGL and investigating the performance and implementation quirks when compared to CUDA.
  4. Testing CUDA algorithism in C or C++.
Commenting is enabled on the zine, so you can add any comments or questions you might have about our work. You can contribute more information to the zine via commenting or sending us a direct email at or