Should a newcomer to writing highly performant code skip CUDA C++ ?
I'm thinking long term about my ML journey and where ML seems to be headed, and I feel that signs keep pointing that I should invest a lot of time in grokking Mojo & MAX.
Perhaps writing some CUDA C++ and reading a lot, potentially to reimplement in Mojo would be useful exercises?
1 Reply
CUDA C is just about enough to learn GPU programming.