A lower-level C-language API for managing computational work in the CUDA platform on NVIDIA GPU hardware.
