Automatically accelerating non-numerical programs by architecture-compiler co-design
Sign Up to like & getrecommendations! Published in 2017 at "Communications of the ACM"
DOI: 10.1145/3139461
Abstract: Because of the high cost of communication between processors, compilers that parallelize loops automatically have been forced to skip a large class of loops that are both critical to performance and rich in latent parallelism.… read more here.
Keywords: automatically accelerating; accelerating non; non numerical; numerical programs ... See more keywords