[Beowulf] scalability of cluster paper
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.
Robert G. Brown rgb at phy.duke.eduFri Feb 18 05:14:39 PST 2005
- Previous message: [Beowulf] scalability of cluster paper
- Next message: [Beowulf] CCGrid 2005: CALL FOR PARTICIPATION
- Messages sorted by: [ date ] [ thread ] [ subject ] [ author ]
On Thu, 17 Feb 2005, Paul Johnson wrote: > I am wondering if anyone could point me to a paper on the scalability > issues of NOWs clusters or > Beowulf clusters using MPI. Im curious what kind of scalability people > see for clusters less than 10 nodes. > Any reference to a paper would be greatly appreciated. I've been doing > alot of scholar.googling but > havent found what Im looking for yet. There is a whole chapter on Amdahl's Law and scaling in my online beowulf book (and several online talks and white papers ditto). Look on http://www.phy.duke.edu/brahma or http://www.phy.duke.edu/~rgb under the Beowulf link. Scalability isn't a question of number of nodes per se, it is a question of the parallel speedup you observe as a FUNCTION of the number of nodes participating in the problem, the problem "size" (really all other parameters, not just size), the network characteristics, and the system CPU/memory/bus characteristics. The presentation I give is a simplified one that lets you see the general idea, but the reality is much more complex for some classes of problem. HTH, rgb -- Robert G. Brown http://www.phy.duke.edu/~rgb/ Duke University Dept. of Physics, Box 90305 Durham, N.C. 27708-0305 Phone: 1-919-660-2567 Fax: 919-660-2525 email:rgb at phy.duke.edu
- Previous message: [Beowulf] scalability of cluster paper
- Next message: [Beowulf] CCGrid 2005: CALL FOR PARTICIPATION
- Messages sorted by: [ date ] [ thread ] [ subject ] [ author ]
More information about the Beowulf mailing list
