Using thread-level speculation to simplify manual parallelization.

Manohar K. Prabhu, Kunle Olukotun

Browse the full PPoPP paper archive.