ABSTRACT
Fetching instructions from a set-associative cache in an embedded processor can consume a large amount of energy due to the tag checks performed. Recent proposals to address this issue involve predicting or memoizing the correct way to access. However, they also require significant hardware storage which negates much of the energy saving.
This paper proposes way-placement to save instruction cache energy. The compiler places the most frequently executed instructions at the start of the binary and at runtime these are mapped to explicit ways within the cache. We compare with a state-of-the-art hardware technique and show that our scheme saves almost 50% of the instruction cache energy compared to 32% for the hardware approach. We report results on a variety of cache sizes and associativities, achieving 59% instruction cache energy savings and an ED product of 0.80 in the best configuration with negligible hardware overhead and no ISA changes.
- N. Bellas et al. Architectural and compiler techniques for energy reduction in high-performance microprocessors. IEEE Transactions on VLSI, 8(3), 2000. Google ScholarDigital Library
- G. Contreras et al. XTREM: a power simulator for the Intel XScale core. In LCTES, 2004. Google ScholarDigital Library
- K. Flautner et al. Drowsy caches: Simple techniques for reducing leakage power. In ISCA-29, 2002. Google ScholarDigital Library
- M. R. Guthaus et al. MiBench: A free, commercially representative embedded benchmark suite. In WWC-4 (MICRO-34), 2001. Google ScholarDigital Library
- C.-H. Hsu et al. Towards efficient supercomputing: A quest for the right metric. In HP-PAC, 2005. Google ScholarDigital Library
- K. Inoue et al. Way-predicting set-associative cache for high performance and low energy consumption. In ISLPED, 1999. Google ScholarDigital Library
- Intel Corporation. Intel XScale microarchitecture. http://www.intel.com/design/intelxscale/.Google Scholar
- T. Ishihara et al. A non-uniform cache architecture for low power system design. In ISLPED, 2005. Google ScholarDigital Library
- C. Jung et al. Instruction cache organisation for embedded low-power processors. IEE Electronics Letters, 37(9), 2001.Google ScholarCross Ref
- S. Kaxiras et al. Cache decay: Exploiting generational behavior to reduce cache leakage power. In ISCA-28, 2001. Google ScholarDigital Library
- J. Kin et al. The filter cache: An energy efficient memory structure. In MICRO-30, 1997. Google ScholarDigital Library
- A. Ma et al. Way memoization to reduce fetch energy in instruction caches. In WCED (ISCA-28), 2001.Google Scholar
- J. Montanaro et al. A 160-MHz, 32-b, 0.5-W CMOS RISC microprocessor. IEEE JSSC, 31(1), 1996.Google Scholar
- R. A. Ravindran et al. Compiler managed dynamic instruction placement in a low-power code cache. In CGO, 2005. Google ScholarDigital Library
- L. Van Put et al. DIABLO: a reliable, retargetable and extensible link-time rewriting framework. In ISSPIT, 2005.Google ScholarCross Ref
- M. Zhang et al. Highly-associative caches for low-power processors. In Koolchips Workshop (MICRO-33), 2000.Google Scholar
Index Terms
- Instruction cache energy saving through compiler way-placement
Recommendations
Compiler-guided next sub-bank prediction for reducing instruction cache leakage energy
Embeded Processors and Systems: Architectural Issues and Solutions for Emerging ApplicationsWith the scaling of technology, leakage energy reduction has become increasingly important, especially for cache memories. Recent studies in drowsy instruction cache show that the leakage energy of the instruction cache can be significantly reduced with ...
An Energy-Efficient Partitioned Instruction Cache Architecture for Embedded Processors
Energy efficiency of cache memories is crucial in designing embedded processors. Reducing energy consumption in the instruction cache is especially important, since the instruction cache consumes a significant portion of total processor energy. This ...
Fast Way-Prediction Instruction Cache for Energy Efficiency and High Performance
NAS '09: Proceedings of the 2009 IEEE International Conference on Networking, Architecture, and StorageThis paper introduces a new way-prediction scheme for achieving low energy consumption and high performance for set-associative instruction cache. The proposed scheme is based on the different cases of how program execution proceeds. By predicting for ...
Comments