Don't have Telegram yet? Try it now!
https://hgpu.org/?p=12842
A Study of the Potential of Locality-Aware Thread Scheduling for GPUs