DMA engine for repeating communication patterns

5,708,779 A 5,761,464 A 5,796,735 A 5,809,278 A 5,825,748 A

1/1998 Graziano etal.

6/1998 Hopkins

8/1998 Miller etal.

9/1998 Watanabe et al.

10/1998 Barleuetal.

(Continued)
OTHER PUBLICATIONS

Definition of "mechanism", Oxford English Dictionary, http://dictionary.oed.com/cgi/entry/00304337? query_Jype=word &queryword=mechanism&first= 1 &max_to_show= 10 &sort_type=alpha&result_place=2&search_id=y2atElGc-l 1603 &hilite+00304337.

(Continued)

Primary Examiner—Henry W. H. Tsai
Assistant Examiner—Ellas Mamo

(74) Attorney, Agent, or Firm—Scully, Scott, Murphy & Presser, PC; Daniel P. Morris, Esq.

A parallel computer system is constructed as a network of interconnected compute nodes to operate a global messagepassing application for performing communications across the network. Each of the compute nodes includes one or more individual processors with memories which run local instances of the global message-passing application operating at each compute node to carry out local processing operations independent of processing operations carried out at other compute nodes. Each compute node also includes a DMA engine constructed to interact with the application via Injection FIFO Metadata describing multiple Injection FIFOs where each Injection FIFO may containing an arbitrary number of message descriptors in order to process messages with a fixed processing overhead irrespective of the number of message descriptors included in the Injection FIFO.

18 Claims, 4 Drawing Sheets

Page 2

U.S. PATENT DOCUMENTS

OTHER PUBLICATIONS

Almasi, et al., "MPI on BlueGene/L: Designing an Efficient General
Purpose Messaging Solution for a Large Cellular System," IBM
Research Report RC22851 (W037-150) Jul. 22, 2003.
Almasi, et al.,"Optimization of MPI Collective Communication on
BlueGene/L Systems," ICS'05, Jun. 20-22, 2005, Boston, MA.
Gara, etal., "Overview of the BlueGene/L system architecture," IBM
J. Res. & Dev., vol. 49, No. 2/3, Mar./May 2005, pp. 195-212.
Huang, et al., "Performance Evaluation of Adaptive MPI,"
PPoPP'06, Mar. 29-31, 2006, New York, New York.
MPI (Message Passing Interface) standards documents, errata, and
archives http://www.mpi-forum.org visited Jun. 16, 2007 (Sections
4.2, 4.4 and 10.4).

David Chaiken, Craig Fields, Kiyoshi Kurihara, Anant Agarwal,
Directory-Based Cache Coherence in Large-Scale Multiprocessors,
Computer, v.23 n.6, p. 49-58, Jun. 1990.

Michel, Dubois, Christoph Scheurich, Faye A. Briggs, Synchronization, Coherence, and Event Ordering in Multiprocessors, Computer, v.21n.2, p. 9-21, Feb. 1988.

Giampapa, et al., "Blue Gene/L advanced diagnostics environment,"

IBM J. Res. & Dev., vol. 49, No. 2/3, Mar./May 2005, pp. 319-331.

IBM Journal of Research and Development, Special Double Issue on

Blue Gene, vol. 49, Nos. 2/3, Mar./May 2005 ("Preface").

IBM Journal of Research and Development, Special Double Issue on

Blue Gene, vol. 49, Nos. 2/3, Mar./May 2005 ("Intro").

"Intel 870: A Building Block For Cost-Effective, Scalable Servers",

Faye Briggs, Michel et al., pp. 36-47, Mar.-Apr. 2002.

Pande, et al. Performance Evaluation and Design Trade-Offs for

Network-On-Chip Interconnect Architectures, 2005, IEEE, pp.

1025-1040.

* cited by examiner