For those of you who may be attending GTC Europe next month in Munich, or who know the schedule:
What highlights you would recommend to people who are interesting in, well, CUDA programming and performance optimization in the plain vanilla sense, rather than something more application-specific?
(I hope this is not off-topic…)
Requests for recommendations are not off-topic here, this isn’t Stackoverflow :-)
Going through the list of sessions, they all seem pretty application specific to me. I also don’t recognize any of the NVIDIA speakers by name.
Some “general, plain vanilla” possibilities:
Multi-GPU programming models
Deep Learning without limits: NVIDIA GPU Cloud
Inside the Volta GPU architecture and CUDA 9
Using OpenMP and OpenACC standards for GPU on Power
Best GPU Code practices for combining OpenACC, CUDA, and OMPSS
Writing elegant host side code … oh, wait … that’s you
It seems I overlooked two thirds of the presentations because I failed to notice that OP’s link is to a tabbed list, rather than a flat one. I’ll work on my reading comprehension.