skip to main content

Title: Collective Memory Transfers for Multi-Core Chips

Future performance improvements for microprocessors have shifted from clock frequency scaling towards increases in on-chip parallelism. Performance improvements for a wide variety of parallel applications require domain-decomposition of data arrays from a contiguous arrangement in memory to a tiled layout for on-chip L1 data caches and scratchpads. How- ever, DRAM performance suffers under the non-streaming access patterns generated by many independent cores. We propose collective memory scheduling (CMS) that actively takes control of collective memory transfers such that requests arrive in a sequential and predictable fashion to the memory controller. CMS uses the hierarchically tiled arrays formal- ism to compactly express collective operations, which greatly improves programmability over conventional prefetch or list- DMA approaches. CMS reduces application execution time by up to 32% and DRAM read power by 2.2×, compared to a baseline DMA architecture such as STI Cell.
; ;
Publication Date:
OSTI Identifier:
Report Number(s):
DOE Contract Number:
Resource Type:
Technical Report
Research Org:
Ernest Orlando Lawrence Berkeley National Laboratory, Berkeley, CA (US)
Sponsoring Org:
USDOE Office of Science (SC)
Country of Publication:
United States
97 MATHEMATICS AND COMPUTING DRAM, access stream, stencils, memory bandwidth, collective transfers