Efficient communication using message prediction for clusters of multiprocessors

Concurrency and Computation: Practice and Experience

Volumn 14, Issue 10, 2002, Pages 859-883

  Afsahi, Ahmad ^a   Dimopoulos, Nikitas J ^b  

^a QUEEN S UNIVERSITY   (Canada)

^b UNIVERSITY OF VICTORIA   (Canada)

Author keywords

Clusters; Communication locality; Message passing interface (MPI); Message prediction; Zero copy

Indexed keywords

BUFFER STORAGE; COMPUTER SOFTWARE; COMPUTER WORKSTATIONS; DATA COMMUNICATION SYSTEMS; INTERFACES (COMPUTER);

INTERPROCESSOR COMMUNICATION;

MULTIPROCESSING SYSTEMS;

EID: 0037173977     PISSN: 15320626     EISSN: None     Source Type: Journal    
DOI: 10.1002/cpe.697     Document Type: Article

References (33)

1. Myrinet: A gigabit-per-second local area network
2. Active messages: A mechanism for integrated communication and computation
3. Efficient layering for high speed communication: The MPI over fast message (FM) experience
4. VMMC-2: Efficient support for reliable, connection-oriented communication
5. U-Net: A user-level network interface for parallel and distributed computing
6. The virtual interface architecture
7. Pin-down cache: A virtual memory management technique for zero-copy communication
8. High-performance local area communication with fast sockets
9. Implementing efficient MPI on LAPI for IBM RS/6000 SP systems: Experiences and performance evaluation
10. Implementation and evaluation of MPI on an SMP cluster
11. Efficient communication using message prediction for cluster of multiprocessors
12. Communication prediction in message-passing systems
13. Hiding communication latency in reconfigurable message-passing environments
14. Using prediction to accelerate coherence protocols
15. A virtual memory mapped network interface for the SHRiMP multicomputer
16. BIP-SMP: High performance message passing over a cluster of commodity SMPs
17. Zero-copy TCP in solaris
18. Fbufs: A high-bandwidth cross-domain transfer facility
19. Buffer-safe and cost-driven communication optimization
20. A unified framework for optimizing communication in data-parallel programs
21. Memory sharing predictor: The key to a speculative coherent DSM
22. Improving CC-NUMA performance using instruction-based prediction
23. Speeding up irregular applications in shared-memory multiprocessors: Memory binding and group prefetching
24. Sequential hardware prefetching in shared-memory multiprocessors
25. Tolerating latency through software-controlled prefetching in shared-memory multiprocessors
26. Parallel spectral transform shallow water model: A runtime-tunable parallel benchmark code
27. Characterization of communication patterns in message-passing parallel scientific application programs
28. Architectural support for reducing communication overhead in multiprocessor interconnection networks
29. Towards a communication characterization methodology for parallel applications