Instructions on building & compiling the code (GPU version)

[C-H-Chien]

HI! Since the links to the documentation are broken, I wonder if by any chance you could provide a brief instruction on how I can build/compile PHCpack, especially its GPU version. Thank you!

[C-H-Chien]

Currently, by just scanning through the repository, I found that:

• GPU-specific code lives under src/GPU with multiple CUDA makefiles (Norms, Polynomials, Newton, Track, TensorCore).
• Legacy integrated GPU trackers are wired in src/Objects/makefile_unix, targets like ts_gpupath_d, phcpy2cpath_d.so.
• Older PHCpy3 docs/scripts include GPU tracker entry points, but the newer src/Python/PHCpy path does not appear to expose those GPU APIs directly.

I am still unclear what the proper way to build/run the GPU version of PHCpack is.
I would thus very much appreciate it if a clear guidance could be provided. Thanks again!