There are enough "problems" involved when working with CUDA for it to deserve its own how-to page. Things that often come up are forward and backward compatibility of different CUDA versions (https://docs.nvidia.com/deploy/cuda-compatibility/). Compatibility between CUDA and GPU is another example.
Ideally, the page has a few compile examples, maybe also with a teeny C code example. The page should also illustrate common error messages, if those exist.
There are enough "problems" involved when working with CUDA for it to deserve its own how-to page. Things that often come up are forward and backward compatibility of different CUDA versions (https://docs.nvidia.com/deploy/cuda-compatibility/). Compatibility between CUDA and GPU is another example.
Ideally, the page has a few compile examples, maybe also with a teeny C code example. The page should also illustrate common error messages, if those exist.
Submitting GPU jobs, etc., should remain at https://wynton.ucsf.edu/hpc/scheduler/gpu.html.