simtk.org

Dear OpenMM developers,
do you plan to publish a guide describing how to extend cuda platform of the OpenMM library?
I found that the guide you published on the OpenMM website documents only the OpenCL platform, but I'm interested in cuda.

Not at present. The CUDA platform code is rather convoluted and quite ugly to describe. I'm hoping we can eventually deprecate it and standardize on the OpenCL platform, rather than having to maintain both of them. But if that turns out not to be possible, I'll give the CUDA platform a thorough refactoring and then document it.

Peter