CUDA C: performance measurement and types of memory
Day: Monday, October 14Time: 2:10 PM - 3:00 PM
Place: Snedecor Hall 2113
Viewers will first learn how record the runtime of a CUDA C kernel. Next, they will learn the differences between local, shared, and global device memory, along with how to utilize each type of storage in CUDA C. I will demonstrate the usage of these types of memory with a CUDA C implementation of the dot product.
Materials
SlidesVideo
Example Code
MakefileMakefile
Makefile