@MISC{Kamruzzaman_load-balancedpipeline, author = {Md Kamruzzaman and Steven Swanson and Dean M. Tullsen}, title = {Load-Balanced Pipeline Parallelism}, year = {} }
Share
OpenURL
Abstract
Accelerating a single thread in current parallel systems remains a challenging problem, because sequential threads do not natu-rally take advantage of the additional cores. Recent work shows that automatic extraction of pipeline parallelism is an effective way to speed up single thread execution. However, two problems re-main challenging – load balancing and inter-thread communica-tion. This work shows new mechanism to exploit pipeline paral-lelism that naturally solves the load balancing and communication problems. This compiler-based technique automatically extracts the pipeline stages and executes them in a data parallel fashion, using token-based chunked synchronization to handle sequential stages. This technique provides linear speedup for several appli-cations, and outperforms prior techniques to exploit pipeline par-allelism by as much as 50%.