A Novel Highly Scalable Clustered Architecture with Partially Distributed Pipeline
Partitioning resources like pipelines and register flies among clusters is proved to be an effective way to improve performance and scalability. However, clustered processors may come at a cost of performance degradation, the major cause of which is the limited data locality arising from lack of available registers and functional units. This paper introduces a highly scalable clustered architecture (HiSCA) to improve the scalability and performance of clustered processors. The pipeline of HiSCA provides high performance through inorder issuing, out-of-order execution and parallel but inorder commitment, while releasing instruction issuing from the heavy burden of dynamic scheduling. Experiment results show that, for a 4-cluster and 8issue configuration. HiSCA can achieve 4.6％ improvement in frequency by 4.6％ with minimal hardware overhead, and an average of 13.3％ performance speedup at the cost of 1.9％ overhead to code size, compared with a traditional clustered processor with nearly the same hardware complexity.
clustered architecture instructiondistribution distributed execution distributed commitment
Hu Chen Shuming Chen
Team 7,Computer School National University of Defense Technology Changsha, China, 410073 Computer School National University of Defense Technology Changsha, China, 410073
国际会议
海口
英文
599-603
2011-07-15（万方平台首次上网日期，不代表论文的发表时间）