SCALING SHARED-BUS MULTIPROCESSORS WITH MULTIPLE BUSES AND SHARED CACHES - A PERFORMANCE STUDY

被引:1
|
作者
BERTONI, J [1 ]
BAER, JL [1 ]
WANG, WH [1 ]
机构
[1] INTEL CORP,ARCHITECTURE DEV LAB,SANTA CLARA,CA 95051
关键词
MULTIPROCESSORS; MULTIPLE BUSES; SHARED CACHES;
D O I
10.1016/0141-9331(92)90002-B
中图分类号
TP3 [计算技术、计算机技术];
学科分类号
0812 ;
摘要
The main limitation of shared-bus multiprocessors is that the common bus tends to be the primary source for contention, and thus imposes a limit on the number of processors in the system. Alternative architectural features are necessary to reduce the memory bandwidth demands and to increase the bus bandwidth. In this paper, we investigate the cost-performance effects of two enhancements: higher bus transaction rates, e.g., through the use of multiple buses, and shared two-level caches. The performance figures are obtained via simulation with loads derived from traces of real applications, some of which show a significant skew in the distribution of memory bank access. A new multiple bus scheme, called multiple interleaved buses, is described and analysed. This scheme is a generalization of previous approaches, and attempts to balance performance and cost trade-offs in a snoopy-cache multiprocessor environment. The results from simulation show that multiple interleaved buses perform almost as well as multiple independent buses, but with simpler and less costly implementation. Furthermore, multiple interleaved buses are shown to deliver much better performance than interleaved buses when the skew of accesses across the interleaves is large. Shared second-level caches have been shown to be very effective in the design space under consideration. Such systems might offer considerable implementation economies with relatively small design cost. We show that depending on the design point in question, bus operation buffers might be useful in shared second level caches by reducing the effects of high skew and greater multiprocessing level. With the presence of these buffers, the uses of shared caches resulted in only a small throughput degradation.
引用
收藏
页码:339 / 350
页数:12
相关论文
共 50 条
  • [41] Enhancing the performance of autoscheduling in Distributed Shared Memory multiprocessors
    Nikolopoulos, DS
    Polychronopoulos, ED
    Papatheodorou, TS
    EURO-PAR '98 PARALLEL PROCESSING, 1998, 1470 : 491 - 501
  • [42] EVALUATING DESIGN CHOICES FOR SHARED BUS MULTIPROCESSORS IN A THROUGHPUT-ORIENTED ENVIRONMENT
    CHIANG, MC
    SOHI, GS
    IEEE TRANSACTIONS ON COMPUTERS, 1992, 41 (03) : 297 - 317
  • [43] PERFORMANCE OF MULTIPLE-BUS INTERCONNECTIONS FOR MULTIPROCESSORS
    YANG, Q
    BHUYAN, LN
    JOURNAL OF PARALLEL AND DISTRIBUTED COMPUTING, 1990, 8 (03) : 267 - 273
  • [44] Network Victim Cache: Leveraging Network-on-Chip for Managing Shared Caches in Chip Multiprocessors
    Wang, Jinglei
    Xue, Yibo
    Wang, Haixia
    Wang, Dongsheng
    PROCEEDINGS OF THE 2009 FOURTH INTERNATIONAL CONFERENCE ON EMBEDDED AND MULTIMEDIA COMPUTING, 2009, : 105 - +
  • [45] Reservation-based media access protocol on multiple-bus distributed shared-memory multiprocessors
    Hwang, I-Shyan
    Proceedings of the National Science Council, Republic of China, Part A: Physical Science and Engineering, 1995, 19 (06):
  • [46] AN EVALUATION OF SET-ASSOCIATIVITY IN 2-LEVEL CACHES FOR SHARED-MEMORY MULTIPROCESSORS
    KRISHNAMOORTHY, S
    CHOUDHARY, A
    LECTURE NOTES IN COMPUTER SCIENCE, 1992, 605 : 751 - 766
  • [47] Prefetch Throttling and Data Pinning for Improving Performance of Shared Caches
    Ozturk, Ozcan
    Son, Seung Woo
    Kandemir, Mahmut
    Karakoy, Mustafa
    INTERNATIONAL CONFERENCE FOR HIGH PERFORMANCE COMPUTING, NETWORKING, STORAGE AND ANALYSIS, 2008, : 384 - +
  • [48] The performance value of shared network caches in clustered multiprocessor workstations
    Bennett, JK
    Fletcher, KE
    Speight, WE
    PROCEEDINGS OF THE 16TH INTERNATIONAL CONFERENCE ON DISTRIBUTED COMPUTING SYSTEMS, 1996, : 64 - 74
  • [49] Matrix multiplication performance on commodity shared-memory multiprocessors
    Tsilikas, G
    Fleury, M
    INTERNATIONAL CONFERENCE ON PARALLEL COMPUTING IN ELECTRICAL ENGINEERING, 2004, : 13 - 18
  • [50] Evaluating synchronization on shared address space multiprocessors: Methodology and performance
    Kumar, S
    Jiang, DM
    Chandra, R
    Singh, JP
    PERFORMANCE EVALUATION REVIEW, SPECIAL ISSUE, VOL 27 NO 1, JUNE 1999: ACM SIGMETRICS '99, PROCEEDINGS - INTERNATIONAL CONFERENCE ON MEASUREMENT AND MODELING OF COMPUTER SYSTEMS, 1999, : 23 - 34