Beowulf: A theorical approach
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.
Douglas Eadline deadline at plogic.comMon Jun 26 08:03:29 PDT 2000
- Previous message: Beowulf: A theorical approach
- Next message: Beowulf: A theorical approach
- Messages sorted by: [ date ] [ thread ] [ subject ] [ author ]
On Mon, 26 Jun 2000, Christoph Wasshuber wrote: > "Robert G. Brown" wrote: > > > It was suggested to me offline that one should consider the AGP bus > > itself as a possible interface, as it apparently has a lot of the > > desired characteristics (and since beowulf nodes typically don't need > > the AGP slot anyway). I'm not an AGP bus expert by any means (it's hard > > enough trying to get PCI specs from the net:-) but this might be worth > > looking into. > > Thinking about it a little longer, this looks to me as not such a bad > idea. Has anybody more information on the AGP bus? Or is there maybe > an AGP bus expert lurking on this list? > My only concern is that AGP has to pass through the North Bridge, which > should increase latency quite a bit. This assumes an AGP interface is present. Some new high end server boards do not seem to have this interface. Doug ------------------------------------------------------------------- Paralogic, Inc. | PEAK | Voice:+610.814.2800 130 Webster Street | PARALLEL | Fax:+610.814.5844 Bethlehem, PA 18015 USA | PERFORMANCE | http://www.plogic.com -------------------------------------------------------------------
- Previous message: Beowulf: A theorical approach
- Next message: Beowulf: A theorical approach
- Messages sorted by: [ date ] [ thread ] [ subject ] [ author ]
More information about the Beowulf mailing list
