The Community for Technology Leaders
Green Image
Issue No. 04 - April (2010 vol. 21)
ISSN: 1045-9219
pp: 494-505
Xavier Martorell , Barcelona Supercomputing Center, Barcelona
Eduard Ayguadé , Barcelona Supercomputing Center, Barcelona
Nikola Vujic , Barcelona Supercomputing Center, Barcelona
Marc Gonzàlez , Technical University of Catalonia, Barcelona
ABSTRACT
Ease of programming is one of the main requirements for the broad acceptance of multicore systems without hardware support for transparent data transfer between local and global memories. Software cache is a robust approach to provide the user with a transparent view of the memory architecture; but this software approach can suffer from poor performance. In this paper, we propose a hierarchical, hybrid software-cache architecture that targets enabling prefetch techniques. Memory accesses are classified at compile time into two classes: high locality and irregular. Our approach then steers the memory references toward one of two specific cache structures optimized for their respective access pattern. The specific cache structures are optimized to enable high-level compiler optimizations to aggressively unroll loops, reorder cache references, and/or transform surrounding loops so as to practically eliminate the software-cache overhead in the innermost loop. The cache design enables automatic prefetch and modulo scheduling transformations. Performance evaluation indicates that optimized software-cache structures combined with the proposed prefetch techniques translate into speedup between 10 and 20 percent. As a result of the proposed technique, we can achieve similar performance on the Cell BE processor as on a modern server-class multicore such as the IBM PowerPC 970MP processor for a set of parallel NAS applications.
INDEX TERMS
Multicore processor, local memories, software cache, prefetch code generation.
CITATION
Xavier Martorell, Eduard Ayguadé, Nikola Vujic, Marc Gonzàlez, "Automatic Prefetch and Modulo Scheduling Transformations for the Cell BE Architecture", IEEE Transactions on Parallel & Distributed Systems, vol. 21, no. , pp. 494-505, April 2010, doi:10.1109/TPDS.2009.97
106 ms
(Ver 3.3 (11022016))