Free launch: optimizing GPU dynamic kernel launches through thread reuse.

Guoyang Chen, Xipeng Shen

Browse the full MICRO paper archive.