Value Compression to Reduce Power in Data Caches

Cache memory represents an important percentage of the total energy consumption of today’s processors. This paper proposes a novel cache design based on data compression to reduce the energy consumed by the data cache. The new scheme stores the same amount of information as a conventional cache but in a smaller physical storage. At the same time, the cache latency is preserved, thus no performance penalty is introduced. The benefits of energy vary from 15.5% to 16%, and the reduction in die area ranges from 23% to 40%.

[1]  Jun Yang,et al.  Energy efficient Frequent Value data Cache design , 2002, 35th Annual IEEE/ACM International Symposium on Microarchitecture, 2002. (MICRO-35). Proceedings..

[2]  Doug Burger,et al.  Evaluating Future Microprocessors: the SimpleScalar Tool Set , 1996 .

[3]  Krste Asanovic,et al.  Dynamic zero compression for cache energy reduction , 2000, MICRO 33.

[4]  Jun Yang,et al.  Frequent Value Locality and Value-Centric Data Cache Design , 2000, ASPLOS.

[5]  Thomas M. Conte,et al.  Exploiting program redundancy to improve performance, cost and power consumption in embedded systems , 2000 .

[6]  John Paul Shen,et al.  The block-based trace cache , 1999, ISCA.

[7]  James E. Smith,et al.  Very low power pipelines using significance compression , 2000, MICRO 33.

[8]  Jun Yang,et al.  Frequent value compression in data caches , 2000, MICRO 33.

[9]  Norman P. Jouppi,et al.  Improving direct-mapped cache performance by the addition of a small fully-associative cache and pre , 1990, ISCA 1990.

[10]  Larry L. Biro,et al.  Power considerations in the design of the Alpha 21264 microprocessor , 1998, Proceedings 1998 Design and Automation Conference. 35th DAC. (Cat. No.98CH36175).

[11]  Margaret Martonosi,et al.  Wattch: a framework for architectural-level power analysis and optimizations , 2000, Proceedings of 27th International Symposium on Computer Architecture (IEEE Cat. No.RS00201).

[12]  Margaret Martonosi,et al.  Dynamically exploiting narrow width operands to improve processor power and performance , 1999, Proceedings Fifth International Symposium on High-Performance Computer Architecture.