Which could be useful for people writing code to run across CPUs, GPUs, etc

GTC  Nvidia said the lines are blurring between the standard C++ and Nv’s CUDA C++ library when it comes to parallel execution of code.…