Slide View : Parallel Programming :: Fall 2019 (original) (raw)

GPU Architecture and CUDA Programming

Previous|Next--- Slide 59 of 80

Back to Lecture Thumbnails

anonymoose

I'm still confused about warps. In some slides, it seems like a warp is similar to an execution context for threads, like on this slide--holds an instruction stream. On others, it seems like it's thread-specific memory. Could someone with a better understanding help reword some of this?

Copyright 2019 Stanford University