Don't have Telegram yet? Try it now!
https://hgpu.org/?p=9647
A Visual Approach to Investigating Shared and Global Memory Behavior of CUDA Kernels