A software architecture for global address space communication on clusters: put/get on fast messages
作者： L.A. Giannini , A.A. Chien
关键词:
摘要: Global address space parallel programming models can be an effective alternative to send/receive style communication, simplifying or code generation and increasing performance for certain application types. Traditionally, global mechanisms have been implemented in hardware order provide the necessary communication responsiveness. However new high cluster messaging systems now allow realized efficiently software. We describe a one sided model that is as software layer on top of Illinois Fast Messages (FM) system. evaluate several different implementation architectures remote agent, characterizing their differing characteristics. Our Put/Get FM achieves peak bandwidths put/get operations 67 MBytes/s, overheads few microseconds, read latencies low 26 microseconds Myrinet connected PC cluster. This was released publicly part HPVM 1.0 August 1997, receiving significant usage. It has used Arrays library also serves back-end target PGI's commercial HPF compiler.
ieeecomputersociety.org
computer.org参考文章(13)
G. Shah, J. Nieplocha, J. Mirza, C. Kim, R. Harrison, R.K. Govindaraju, K. Gildea, P. DiNicola, C. Bender, Performance and experience with LAPI-a new high-performance communication library for the IBM RS/6000 SP merged international parallel processing symposium and symposium on parallel and distributed processing. pp. 260- 266 ,(1998) , 10.1109/IPPS.1998.669923
Kenichi Hayashi, Tsunehisa Doi, Takeshi Horie, Yoichi Koyanagi, Osamu Shiraki, Nobutaka Imamura, Toshiyuki Shimizu, Hiroaki Ishihata, Tatsuya Shindo, AP1000+: architectural support of PUT/GET interface for parallelizing compiler architectural support for programming languages and operating systems. ,vol. 28, pp. 196- 207 ,(1994) , 10.1145/195473.195538
Remzi H. Arpaci, David E. Culler, Arvind Krishnamurthy, Steve G. Steinberg, Katherine Yelick, Empirical evaluation of the CRAY-T3D: a compiler perspective international symposium on computer architecture. ,vol. 23, pp. 320- 331 ,(1995) , 10.1145/223982.224443
S. Pakin, V. Karamcheti, A.A. Chien, Fast messages: efficient, portable communication for workstation clusters and MPPs IEEE Concurrency. ,vol. 5, pp. 60- 73 ,(1997) , 10.1109/4434.588295
Arvind Krishnamurthy, Klaus E. Schauser, Chris J. Scheiman, Randolph Y. Wang, David E. Culler, Katherine Yelick, Evaluation of architectural support for global address-based communication in large-scale parallel machines Proceedings of the seventh international conference on Architectural support for programming languages and operating systems - ASPLOS-VII. ,vol. 31, pp. 37- 48 ,(1996) , 10.1145/237090.237147
N.J. Boden, D. Cohen, R.E. Felderman, A.E. Kulawik, C.L. Seitz, J.N. Seizovic, Wen-King Su, Myrinet: a gigabit-per-second local area network IEEE Micro. ,vol. 15, pp. 29- 36 ,(1995) , 10.1109/40.342015
B.-H. Lim, P. Heidelberger, P. Pattnaik, M. Snir, Message proxies for efficient, protected communication on SMP clusters high-performance computer architecture. pp. 116- 127 ,(1997) , 10.1109/HPCA.1997.569647
J. Nieplocha, R.J. Harrison, Shared memory NUMA programming on I-WAY high performance distributed computing. pp. 432- 441 ,(1996) , 10.1109/HPDC.1996.546214
R. Horst, Tnet: a reliable system area network for 1/0 and ipc Symposium Record Hot Interconnects II. pp. 145- 150 ,(1994) , 10.1109/CONECT.1994.765344
Jaroslaw Nieplocha, Richard J. Littlefield, Robert J. Harrison, Global Arrays: a portable "shared-memory" programming model for distributed memory computers conference on high performance computing (supercomputing). pp. 340- 349 ,(1994) , 10.5555/602770.602833