Using Magnetic RAM to Build Low-Power and Soft Error-Resilient L1 Cache

被引:22
作者
Sun, Hongbin [1 ]
Liu, Chuanyin [1 ]
Xu, Wei [2 ]
Zhao, Jizhong [1 ]
Zheng, Nanning [1 ]
Zhang, Tong [3 ]
机构
[1] Xi An Jiao Tong Univ, Xian 710049, Shaanxi, Peoples R China
[2] Marvell Technol, Santa Clara, CA 95054 USA
[3] Rensselaer Polytech Inst, Elect Comp & Syst Engn Dept, Troy, NY 12180 USA
基金
中国国家自然科学基金; 美国国家科学基金会;
关键词
Cache; low power; magnetic RAM (MRAM); reliability; soft error; MEMORY; ARCHITECTURE; MRAM;
D O I
10.1109/TVLSI.2010.2090914
中图分类号
TP3 [计算技术、计算机技术];
学科分类号
0812 ;
摘要
Due to its great scalability, fast read access, low leakage power, and nonvolatility, magnetic random access memory (MRAM) appears to be a promising memory technology for on-chip cache memory in microprocessors. However, the write-to-MRAM process is relatively slow and results in high dynamic power consumption. Such inherent disadvantages of MRAM make researchers easily conclude that MRAM can only be used for low-level caches (e. g., L2 or L3 cache), where cache memories are less frequently accessed and slow write to MRAM can be more easily compensated using simple architectural techniques. By developing a hybrid cache architecture, this paper attempts to show that, with appropriate architecture design, MRAM can also be used in L1 cache to improve both the energy efficiency and soft error immunity. The basic idea is to supplement the MRAM L1 cache with several small SRAM buffers, which can substantially mitigate the performance degradation and dynamic energy overhead induced by MRAM write operations. Moreover, the proposed hybrid cache architecture is also an efficient solution to protect cache memory from radiation-induced soft errors, as MRAM is inherently invulnerable to emissive particles. Simulation results show that, with only less than 2% performance degradation, the proposed design approach can reduce the power consumption by up to 76.1% on average compared with the traditional SRAM L1 cache. In addition, the architectural vulnerability factor of L1 data cache is reduced from 28.3% to as low as 0.5%.
引用
收藏
页码:19 / 28
页数:10
相关论文
共 24 条
  • [1] [Anonymous], P INT WORKSH ARCH SU
  • [2] [Anonymous], CACTI: An Integrated Cache and Memory Access Time, Cycle Time, Area, Leakage, and Dynamic Power Model
  • [3] Baumann R. C., 2001, IEEE Transactions on Device and Materials Reliability, V1, P17, DOI 10.1109/7298.946456
  • [4] Computing architectural vulnerability factors for address-based structures
    Biswas, A
    Racunas, P
    Cheveresan, R
    Emer, J
    Mukherjee, SS
    Rangan, R
    [J]. 32ND INTERNATIONAL SYMPOSIUM ON COMPUTER ARCHITECTURE, PROCEEDINGS, 2005, : 532 - 543
  • [5] ERROR-CORRECTING CODES FOR SEMICONDUCTOR MEMORY APPLICATIONS - A STATE-OF-THE-ART REVIEW
    CHEN, CL
    HSIAO, MY
    [J]. IBM JOURNAL OF RESEARCH AND DEVELOPMENT, 1984, 28 (02) : 124 - 134
  • [6] Dong XY, 2008, DES AUT CON, P554
  • [7] A comparative study on the soft-error rate of flip-flops from 90-nm production libraries
    Heijmen, Tino
    Roche, Philippe
    Gasiot, Gilles
    Forbes, Keith R.
    [J]. 2006 IEEE INTERNATIONAL RELIABILITY PHYSICS SYMPOSIUM PROCEEDINGS - 44TH ANNUAL, 2006, : 204 - +
  • [8] Hennessy J.L., 2006, Computer Architecture: A Quantitative Approach'', V4th
  • [9] Itoh K, 2002, INT CONF MICROELECTR, P497, DOI 10.1109/MIEL.2002.1003306
  • [10] Jouppi N. P., 1990, Proceedings. The 17th Annual International Symposium on Computer Architecture (Cat. No.90CH2887-8), P364, DOI 10.1109/ISCA.1990.134547