[Beowulf] Re: torus versus (fat) tree topologies
Many of your questions may have already been answered in earlier discussions or in the FAQ. The search results page will indicate current discussions as well as past list serves, articles, and papers.
Dan Kidger daniel.kidger at quadrics.comSun Nov 21 04:50:56 PST 2004
- Previous message: [Beowulf] Re: torus versus (fat) tree topologies
- Next message: [Beowulf] Re: torus versus (fat) tree topologies
- Messages sorted by: [ date ] [ thread ] [ subject ] [ author ]
On Saturday 20 November 2004 9:53 pm, Pfenniger Daniel wrote: > > I could not however easily find a complete list of Infiniband price > > anywhere. Most all vendors I tried require requesting a quote. Does > >anyone have a nice online source for such pricing? > > Googling gave a German site with clear prices: > http://www.olmos.de/infini.html?lang=en&tax=no&xtra=¤cy=USD > > The cost of 24 port switches has decreased by more than a factor 10 > with repect to list prices of a defunct InfiniBand company 20 months ago, > without taking the falling USD into account. I believe that currently Myrinet (revE) is more expensive than Quadrics, but I thought that Infiniband was now often cheaper than Quadrics. However this German website seems to contradict this view? Can someone on the list who understands Voltaire's products, say how much an Infinband switch for a 128node cluster would cost? It is an ISR-9288 ($29,950) with four sLB-24 cards ($10,950 each) = $73,750 or am I missing something ? As for the other components; Voltaire, Quadrics and Myricom(revE) cards all cost pretty much the same. cables costs are almost identical (remembering revE uses 2 cables per node). All three vendors have free open-source drivers, libraries and MPI. Daniel. -------------------------------------------------------------- Dr. Dan Kidger, Quadrics Ltd. daniel.kidger at quadrics.com One Bridewell St., Bristol, BS1 2AA, UK 0117 915 5505 ----------------------- www.quadrics.com --------------------
- Previous message: [Beowulf] Re: torus versus (fat) tree topologies
- Next message: [Beowulf] Re: torus versus (fat) tree topologies
- Messages sorted by: [ date ] [ thread ] [ subject ] [ author ]
More information about the Beowulf mailing list
