My point was not that the application is naive. It's that POWER5+ clock-per-clock gets outdone by a Pentium-M on this workload.
Hence, "on a processor to processor basis." Of course total throughput of the IBM is gonna be better than the Dell notebook (and it had better be given the 100X price difference). But that's pretty obvious because the IBM has lots of cores. If you can have lots of cores (the test machine in question has 8 cores) you can get an HP DL585 with four AMD Opteron 8220 SE's for a total of 8 cores @ 2.8GHz, for much less money, and that would fly too. What's sauce for the goose is sauce for the gander. On 2/28/07, thad <[EMAIL PROTECTED]> wrote:
thats what i also suspect when dba and application development people complains their apps is performing slow on multiple cpus. i told them to verify their apps if it know how to thread on multi-cores or smp. On 2/28/07, Orlando Andico <[EMAIL PROTECTED]> wrote: > Obviously it does not use multi cores. That's wlhy I said "on a > processor to processor basis." The application is single-threaded.
_________________________________________________ Philippine Linux Users' Group (PLUG) Mailing List [email protected] (#PLUG @ irc.free.net.ph) Read the Guidelines: http://linux.org.ph/lists Searchable Archives: http://archives.free.net.ph
