Performance of message-passing systems using a zero-copy communication protocol

被引:1
|
作者
Melas, P [1 ]
Zaluska, EJ [1 ]
机构
[1] Univ Southampton, Southampton SO9 5NH, Hants, England
来源
1998 INTERNATIONAL CONFERENCE ON PARALLEL ARCHITECTURES AND COMPILATION TECHNIQUES, PROCEEDINGS | 1998年
关键词
D O I
10.1109/PACT.1998.727260
中图分类号
TP3 [计算技术、计算机技术];
学科分类号
0812 ;
摘要
Despite technological advantages in microprocessors and network technology over the last few years, commercially-available networks of workstations (NOWs) contain inherent communication bottlenecks Traditional layered network protocols will inevitably fail to achieve high throughput if they access data several times. As a result, applications an NOWs often fail to observe the performance speed-up that might be expected. Network protocols which avoid routing through the kernel can remove this limit an communication performance and support very high transmission speeds, turning NOWs into an attractive alternative to Massively-Parallel Processors. This paper investigates the improvements in performance of an entire NOW system that are possible by changing communication protocols. A commodity workstation cluster with a zero-copy communication protocol (Basic Interface Parallelism) is examined and compared with both conventional protocol implementations and parallel systems in order to understand and assess the improvements in performance achieved.
引用
收藏
页码:264 / 271
页数:8
相关论文
共 50 条
  • [11] Communication prediction in message-passing multiprocessors
    Afsahi, A
    Dimopoulos, NJ
    HIGH PERFORMANCE COMPUTING SYSTEMS AND APPLICATIONS, 2002, 657 : 253 - 271
  • [12] LyraNET: A zero-copy TCP/IP protocol stack for embedded systems
    Mei-Ling Chiang
    Yun-Chen Li
    Real-Time Systems, 2006, 34 : 5 - 18
  • [13] A performance instrumentation framework to characterize computation-communication overlap in message-passing systems
    Shet, Aniruddha G.
    Sadayappan, P.
    Bernholdt, David E.
    Nieplocha, Jarek
    Tipparaju, Vinod
    2006 IEEE INTERNATIONAL CONFERENCE ON CLUSTER COMPUTING, VOLS 1 AND 2, 2006, : 455 - +
  • [14] LyraNET: A zero-copy TCP/IP protocol stack for embedded systems
    Chiang, Mei-Ling
    Li, Yun-Chen
    REAL-TIME SYSTEMS, 2006, 34 (01) : 5 - 18
  • [15] Implementation of a simultaneous message-passing protocol using optical vortices
    Szatkowski, Mateusz
    Koechlin, Julian
    Lopez-Mago, Dorilian
    OPTICS AND LASER TECHNOLOGY, 2021, 133
  • [16] A New Approach to Zero-Copy Message Passing with Reversible Memory Allocation in Multi-Core Architectures
    Swenson, Brian Paul
    Riley, George F.
    2012 ACM/IEEE/SCS 26TH WORKSHOP ON PRINCIPLES OF ADVANCED AND DISTRIBUTED SIMULATION (PADS), 2012, : 44 - 52
  • [17] RDMA Managed Buffers: A Case for Accelerating Communication Bound Processes via Fine-Grained Events for Zero-Copy Message Passing
    Wickramasinghe, Udayanga
    Lumsdaine, Andrew
    Ekanayake, Saliya
    Swany, Martin
    2019 18TH INTERNATIONAL SYMPOSIUM ON PARALLEL AND DISTRIBUTED COMPUTING (ISPDC 2019), 2019, : 121 - 130
  • [18] A message-passing solver for linear systems
    Shental, Ori
    Bickson, Danny
    Siegel, Paul H.
    Wolf, Jack K.
    Dolev, Danny
    2008 INFORMATION THEORY AND APPLICATIONS WORKSHOP, 2008, : 403 - 403
  • [19] Performance Simulation of Non-blocking Communication in Message-Passing Applications
    Boehme, David
    Hermanns, Marc-Andre
    Geimer, Markus
    Wolf, Felix
    EURO-PAR 2009 PARALLEL PROCESSING WORKSHOPS, 2010, 6043 : 208 - 217
  • [20] Performance of message-passing MATLAB toolboxes
    Fernández, J
    Cañas, A
    Díaz, AF
    González, J
    Ortega, J
    Prieto, A
    HIGH PERFORMANCE COMPUTING FOR COMPUTATIONAL SCIENCE - VECPAR 2002, 2003, 2565 : 228 - 241