GigaBit Performance under NT
The recent interest and growing popularity of commodity-based cluster computing has created a demand for low-latency, high-bandwidth interconnect technologies. Early cluster systems have used expensive but fast interconnects such as Myrinet or SCI. Even though these technologies provide low-latency, high-bandwidth communications, the cost of an interface card almost matches that of individual computers in the cluster. Even though these specialist technologies are popular, there is a growing demand for Ethernet which can provide a low-risk and upgradeable path with which to link clusters together. In this paper we compare and contrast the low-level performance of a range of Giganet network cards under Windows NT using MPI and PVM. In the first part of the paper we discuss our motivation and rationale for undertaking this work. We then move on to discuss the systems that we are using and our methods for assessing these technologies. In the second half of the paper we present our results and discuss our findings. In the final section of the paper we summarize our experiences and then briefly mention further work we intend to undertake.
Keywordscluster interconnect communication network Gigabit Ethernet PVM MPI performance evaluation
Unable to display preview. Download preview PDF.
- 1.A. Geist, Cluster Computing: The Wave of the future, Springer Verlag Lecture Notes in Computer Science, May 1994.Google Scholar
- 2.The PVM project-http://www.epm.ornl.gov/pvm/
- 3.MPI Forum-http://www.mpi-forum.org/docs/docs.html
- 4.Message Passing Interface Forum, MPI: A Message-Passing Interface Standard, University of Tennessee, Knoxville, Report No. CS-94-230, May 5, 1994Google Scholar
- 6.W. Gropp, et. al., A high-performance, portable implementation of the MPI message passing interface standard-http://www-c.mcs.anl.gov/mpi/mpicharticle/paper.html
- 7.W. Gropp and B. Smith, Chameleon parallel programming tools users manual. Technical Report ANL-93/23, Argonne National Laboratory, March 1993.Google Scholar
- 8.PVM: A Users’ Guide and Tutorial For Networked Parallel Computing-http://www.netlib.org/pvm3/book/pvm-book.html
- 9.Gigabit Ethernet Alliance-Gigabit Ethernet: Accelerating the standard for speed, http://www.gigabit-ethernet.org/technology/whitepapers, September 1999.
- 10.Ethernet Segment Limits.-http://www.gigabit-ethernet.org/technology/
- 12.MPI Software Technology, Inc. — http://www.mpi-softtech.com/
- 14.VIA — http://www.viaarch.com
- 16.WINdows based PARallel computing-http://www.genias.de/
- 18.R. Buttler and E. Lusk, User’s Guide to the p4 Parallel Programming System, ANL-92/17, Argonne National Laboratory, October 1992.Google Scholar
- 21.Packet Engine-http://www.packetengines.com/index4.html
- 23.N. Boden, et. al. Myrinet — A Gbps LAN. IEEE Micro, Vol. 15, No.1, February 1995. http://www.myri.com/
- 24.Dolphin Interconnect Solutions-http://www.dolphinics.no/