Lessons Loearned from Implementing BSP

摘要

We focus on two criticisms of Bulk Synchronous Parallelism (BSP): that delaying communication until specific points in a program causes poor performance, and that frequent barrier synchronistations are too expensive for high-performance parallel computing. We show that these criticisms are misguided, not just about BSP but about parallel progrmaming in general, because they are based on misconceptions about the origins of poor performance. The main implication for parallel programming is that higher levels of abstraction do not only make software construction easier-they also make high-performance implementation easier.