SwePub
Sök i LIBRIS databas

  Utökad sökning

WFRF:(Fechner L.)
 

Sökning: WFRF:(Fechner L.) > Software news and u...

LIBRIS Formathandbok  (Information om MARC21)
FältnamnIndikatorerMetadata
00003613naa a2200409 4500
001oai:DiVA.org:kth-82625
003SwePub
008120212s2007 | |||||||||||000 ||eng|
0247 a https://urn.kb.se/resolve?urn=urn:nbn:se:kth:diva-826252 URI
0247 a https://doi.org/10.1002/jcc.207032 DOI
040 a (SwePub)kth
041 a engb eng
042 9 SwePub
072 7a ref2 swepub-contenttype
072 7a art2 swepub-publicationtype
1001 a Kutzner, Carstenu Max-Planck Institut Göttingen4 aut
2451 0a Software news and update :b Speeding up parallel GROMACS on high-latency networks
264 c 2007
264 1b Wiley,c 2007
338 a print2 rdacarrier
500 a QC 20120302
520 a We investigate the parallel scaling of the GROMACS molecular dynamics code on Ethernet Beowulf clusters and what prerequisites are necessary for decent scaling even on such clusters with only limited bandwidth and high latency. GROMACS 3.3 scales well on supercomputers like the IBM p690 (Regatta) and on Linux clusters with a special interconnect like Myrinet or Infiniband. Because of the high single-node performance of GROMACS, however, on the widely used Ethernet switched clusters, the scaling typically breaks down when more than two computer nodes are involved, limiting the absolute speedup that can be gained to about 3 relative to a single-CPU run. With the LAM MPI implementation, the main scaling bottleneck is here identified to be the all-to-all communication which is required every time step. During such an all-to-all communication step, a huge amount of messages floods the network, and as a result many TCP packets are lost. We show that Ethernet flow control prevents network congestion and leads to substantial scaling improvements. For 16 CPUs, e.g., a speedup of 11 has been achieved. However, for more nodes this mechanism also fails. Having optimized an all-to-all routine, which sends the data in an ordered fashion, we show that it is possible to completely prevent packet loss for any number of multi-CPU nodes. Thus, the GROMACS scaling dramatically improves, even for switches that lack flow control. In addition, for the common HP ProCurve 2848 switch we find that for optimum all-to-all performance it is essential how the nodes are connected to the switch's ports. This is also demonstrated for the example of the Car-Parinello MD code.
650 7a NATURVETENSKAPx Kemix Teoretisk kemi0 (SwePub)104072 hsv//swe
650 7a NATURAL SCIENCESx Chemical Sciencesx Theoretical Chemistry0 (SwePub)104072 hsv//eng
650 7a NATURVETENSKAPx Data- och informationsvetenskapx Programvaruteknik0 (SwePub)102052 hsv//swe
650 7a NATURAL SCIENCESx Computer and Information Sciencesx Software Engineering0 (SwePub)102052 hsv//eng
7001 a van der Spoel, Davidu Uppsala University4 aut
7001 a Fechner, Martinu Max-Planck Institut Göttingen4 aut
7001 a Lindahl, Erik,d 1972-u Stockholm University4 aut0 (Swepub:kth)u1u9f2s7
7001 a Schmitt, Udo Wu Max-Planck Institut Göttingen4 aut
7001 a de Groot, Bert Lu Max-Planck Institut Göttingen4 aut
7001 a Grubmüller, Helmutu Max-Planck Institut Göttingen4 aut
7102 a Max-Planck Institut Göttingenb Uppsala University4 org
7730 t Journal of Computational Chemistryd : Wileyg 28:12, s. 2075-84q 28:12<2075-84x 0192-8651x 1096-987X
8564 u https://onlinelibrary.wiley.com/doi/pdfdirect/10.1002/jcc.20703
8564 8u https://urn.kb.se/resolve?urn=urn:nbn:se:kth:diva-82625
8564 8u https://doi.org/10.1002/jcc.20703

Hitta via bibliotek

Till lärosätets databas

Kungliga biblioteket hanterar dina personuppgifter i enlighet med EU:s dataskyddsförordning (2018), GDPR. Läs mer om hur det funkar här.
Så här hanterar KB dina uppgifter vid användning av denna tjänst.

 
pil uppåt Stäng

Kopiera och spara länken för att återkomma till aktuell vy