JETTY: filtering snoops for reduced energy consumption in SMP servers
- 13 November 2002
- conference paper
- Published by Institute of Electrical and Electronics Engineers (IEEE)
- No. 15300897,p. 85-96
- https://doi.org/10.1109/hpca.2001.903254
Abstract
We propose methods for reducing the energy consumed by snoop requests in snoopy bus-based symmetric multiprocessor (SMP) systems. Observing that a large fraction of snoops do not find copies in many of the other caches, we introduce JETTY, a small, cache-like structure. A JETTY is introduced in-between the bus and the L2 backside of each processor. There it filters the vast majority of snoops that would not find a locally cached copy. Energy is reduced as accesses to the much more energy demanding L2 tag arrays are decreased. No changes in the existing coherence protocol are required and no performance loss is experienced. We evaluate our method on a 4-way SMP server using a set of shared-memory applications. We demonstrate that a very small JETTY filters 74% (average) of all snoop-induced tag accesses that would miss. This results in an average energy reduction of 29% (range: 12% to 40%) measured as a fraction of the energy required by all L2 accesses (both tag and data arrays).Keywords
This publication has 20 references indexed in Scilit:
- Wisconsin Wind Tunnel II: a fast, portable parallel architecture simulatorIEEE Concurrency, 2000
- Gated-V/sub dd/: a circuit technique to reduce leakage in deep-submicron cache memoriesPublished by Institute of Electrical and Electronics Engineers (IEEE) ,2000
- Memory sharing predictorACM SIGARCH Computer Architecture News, 1999
- Performance of database workloads on shared-memory systems with out-of-order processorsPublished by Association for Computing Machinery (ACM) ,1998
- A 250-MHz 5-W PowerPC microprocessor with on-chip L2 cache controllerIEEE Journal of Solid-State Circuits, 1997
- Increasing cache port efficiency for dynamic superscalar microprocessorsPublished by Association for Computing Machinery (ACM) ,1996
- Cache design trade-offs for power and performance optimizationPublished by Association for Computing Machinery (ACM) ,1995
- Reducing the frequency of tag compares for low power I-cache designPublished by Association for Computing Machinery (ACM) ,1995
- Using processor-cache affinity information in shared-memory multiprocessor schedulingIEEE Transactions on Parallel and Distributed Systems, 1993
- Analysis of cache invalidation patterns in multiprocessorsPublished by Association for Computing Machinery (ACM) ,1989