The Community for Technology Leaders
Proceedings of the 22nd International Conference on Parallel Architectures and Compilation Techniques (2007)
Brasov, Romania
Sept. 15, 2007 to Sept. 19, 2007
ISSN: 1089-795X
ISBN: 0-7695-2944-5
pp: 175-188
Trevor Mudge , University of Michigan-Ann Arbor, USA
Ronald G. Dreslinkski , University of Michigan-Ann Arbor, USA
Bo Zhai , University of Michigan-Ann Arbor, USA
David Blaauw , University of Michigan-Ann Arbor, USA
Dennis Sylvester , University of Michigan-Ann Arbor, USA
ABSTRACT
Subthreshold circuit design, while energy efficient, has the drawback of performance degradation. To retain the excellent energy efficiency while reducing performance loss, we propose to investigate near subthreshold techniques on chip multiprocessors (CMP). We show that logic and memory cells have different optimal supply and threshold voltages, therefore we propose to allow the cores and memory to operate in different voltage regions. With the memory operating at a different voltage, we then explore the design space in which several slower cores clustered together share a faster L1 cache. We show that an architecture such as this is optimal for energy efficiency. In particular, SPLASH2 benchmarks show a 53% energy reduction over the conventional CMP approach (70% energy reduction over a single core machine). In addition we explore the design trade-offs that occur if we have a separate instruction and data cache. We show that some applications prefer the data cache to be clustered while the instruction cache is kept private to the core allowing further energy savings of a 77% reduction over a single core machine.
INDEX TERMS
null
CITATION
Trevor Mudge, Ronald G. Dreslinkski, Bo Zhai, David Blaauw, Dennis Sylvester, "An Energy Efficient Parallel Architecture Using Near Threshold Operation", Proceedings of the 22nd International Conference on Parallel Architectures and Compilation Techniques, vol. 00, no. , pp. 175-188, 2007, doi:10.1109/PACT.2007.14
91 ms
(Ver 3.3 (11022016))