A Shared-Way Set Associative architecture for on-chip caches

被引:0
|
作者
Hamkalo, JL [1 ]
Djordjalian, A [1 ]
Cernuschi-Frías, B [1 ]
机构
[1] Univ Buenos Aires, Fac Ingn, RA-1063 Buenos Aires, DF, Argentina
关键词
cache memory; associativity;
D O I
暂无
中图分类号
TP18 [人工智能理论];
学科分类号
081104 ; 0812 ; 0835 ; 1405 ;
摘要
The scheme described in this paper is an alternative cache memory organization called "Shared-Way Set Associative" (SWSA). It consists of a modified two-way set associative scheme in which one bank is larger than the other. We show how better use of memory is obtained, without the costs that higher-associativities have. An expression for calculating the non-integer degree of associativity of SWSA caches is given. Several replacement policies are discussed. Miss rate statistics for the SPEC95 benchmarks are presented for first and second level SWSA caches, together with a detailed analysis of conflicts using the D3C classification of misses. For large caches the miss rates of SWSA caches are similar to those of 33% larger two-way caches. The issue of access time is addressed, and we explain why SWSA caches may have advantages, specially for configurations with very unbalanced banks which have miss rates that are very similar to those of slightly smaller two-way caches.
引用
收藏
页码:125 / 128
页数:4
相关论文
共 50 条
  • [41] Resilience Analysis: Tightening the CRPD bound for set-associative caches
    Altmeyer, Sebastian
    Maiza , Claire
    Reineke, Jan
    [J]. ACM SIGPLAN NOTICES, 2010, 45 (04) : 153 - 162
  • [42] Achieving predictable performance with on-chip shared L2 caches for manycore-based real-time systems
    Cho, Sangyeun
    Jin, Lei
    Lee, Kiyeon
    [J]. 13TH IEEE INTERNATIONAL CONFERENCE ON EMBEDDED AND REAL-TIME COMPUTING SYSTEMS AND APPLICATIONS, PROCEEDINGS, 2007, : 3 - +
  • [43] CCC: Crossbar connected caches for reducing energy consumption of on-chip multiprocessors
    Li, L
    Vijaykrishnan, N
    Kandemir, M
    Irwin, MJ
    Kadayif, I
    [J]. EUROMICRO SYMPOSIUM ON DIGITAL SYSTEM DESIGN, PROCEEDINGS, 2003, : 41 - 48
  • [44] Maximizing Energy Efficiency of On-Chip Caches Exploiting Hybrid Memory Structure
    Xu, Hongjie
    Shiomi, Jun
    Ishihara, Tohru
    Onodera, Hidetoshi
    [J]. 2018 28TH INTERNATIONAL SYMPOSIUM ON POWER AND TIMING MODELING, OPTIMIZATION AND SIMULATION (PATMOS), 2018, : 237 - 242
  • [45] Performance aware on-chip communication synthesis and optimization for shared multi-bus based architecture
    Pandey, S
    Glesner, M
    Mühlhäuser, M
    [J]. SBCCI 2005: 18th Symposium on Integrated Circuits and Systems Design, Proceedings, 2005, : 230 - 235
  • [46] Nonuniform cache architectures for wire-delay dominated on-chip caches
    Kim, C
    Burger, D
    Keckler, SW
    [J]. IEEE MICRO, 2003, 23 (06) : 99 - 107
  • [47] An architecture for high-performance scalable shared-memory multiprocessors exploiting on-chip integration
    Acacio, ME
    González, J
    García, JM
    Duato, J
    [J]. IEEE TRANSACTIONS ON PARALLEL AND DISTRIBUTED SYSTEMS, 2004, 15 (08) : 755 - 768
  • [48] MemCAM: A Hybrid Memristor-CMOS CAM Cell for On-Chip Caches
    Sadiq, Zareen
    Hasan, Shehzad
    [J]. IEEE ACCESS, 2021, 9 : 21296 - 21305
  • [49] Subchannel Scheduling for Shared Optical On-chip Buses
    Werner, Sebastian
    Navaridas, Javier
    Lujan, Mikel
    [J]. 2017 IEEE 25TH ANNUAL SYMPOSIUM ON HIGH-PERFORMANCE INTERCONNECTS (HOTI), 2017, : 49 - 56
  • [50] ARCHITECTURE FOR AN ISDN CHIP SET
    CROOKS, P
    [J]. ELECTRONIC ENGINEERING, 1990, 62 (765): : 27 - &