[Beowulf] x86-64 NUMA vs SMP kernel: appl. performance?
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 Sep 24 12:04:05 PDT 2004
- Previous message: [Beowulf] x86-64 NUMA vs SMP kernel: appl. performance?
- Next message: [Beowulf] x86-64 NUMA vs SMP kernel: appl. performance?
- Messages sorted by: [ date ] [ thread ] [ subject ] [ author ]
On Fri, 24 Sep 2004, Greg Lindahl wrote: > > I have possible choice between using of SMP or NUMA-enabled x86_64 > > kernels (2.4.21 from SuSE Linux 9.0 distributive). > > Mikhail, > > If I wanted to make a single generalization about Opteron kernels, it > would be "Use a 2.6 kernel!" However, we've found that picking the > right compiler is an even larger effect. > > -- greg I'd second the bit about the 2.6 kernel. In fact, I'd say that the 2.4 kernel is almost broken on x64. Almost because it doesn't usually actually not work, kinda sorta. It just works terribly poorly. I'm interested in kernel effects on top of using e.g. FC2 with a 2.6.x kernel. I see balanced CPU (which one didn't in 2.4) and good multitasking (ditto) and a fairly major improvement in throughput (all the CPU you lost one the off processor). What compilers have you tried, and what improvements do they produce? 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] x86-64 NUMA vs SMP kernel: appl. performance?
- Next message: [Beowulf] x86-64 NUMA vs SMP kernel: appl. performance?
- Messages sorted by: [ date ] [ thread ] [ subject ] [ author ]
More information about the Beowulf mailing list
