Most notably, the chipmaker announced a compiler source code enabling software developers to add new languages and architecture support to Nvidia’s CUDA parallel programming model. The new ...