ARCc: A case for an architecturally redundant cache-coherence architecture for large multicores
Authors:
Open Access
- 1 October 2011
- conference paper
- Published by Institute of Electrical and Electronics Engineers (IEEE)
- p. 411-418
Abstract
This paper proposes an architecturally redundant cache-coherence architecture (ARCc) that combines the directory and shared-NUCA based coherence protocols to improve performance, energy and dependability. Both coherence mechanisms co-exist in the hardware and ARCc enables seamless transition between the two protocols. We present an online analytical model implemented in the hardware that predicts performance and triggers a transition between the two coherence protocols at application-level granularity. The ARCc architecture delivers up to 1.6× higher performance and up to 1.5× lower energy consumption compared to the directory-based counterpart. It does so by identifying applications which benefit from the large shared cache capacity of shared-NUCA because of lower off-chip accesses, or where remote-cache word accesses are efficient.Keywords
This publication has 15 references indexed in Scilit:
- Analyzing Parallel Programs with PinComputer, 2010
- A 45nm 8-core enterprise Xeon® processor2009
- An 80-Tile Sub-100-W TeraFLOPS Processor in 65-nm CMOSIEEE Journal of Solid-State Circuits, 2008
- Thousand core chipsProceedings of the 39th conference on Design automation - DAC '02, 2007
- Designing Reliable Systems from Unreliable Components: The Challenges of Transistor Variability and DegradationIEEE Micro, 2005
- The SPLASH-2 programs: characterization and methodological considerations2002
- Using simple page placement policies to reduce the cost of cache fills in coherent shared-memory systems2002
- An adaptive, non-uniform cache structure for wire-delay dominated on-chip caches2002
- Operating system support for improving data locality on CC-NUMA compute serversACM SIGPLAN Notices, 1996
- Directory-based cache coherence in large-scale multiprocessorsComputer, 1990