CPW benchmark is very close to TPC-C. The major difference is that TPC-C requires independent auditing, and is a very expensive undertaking. I've had a chance before to mention that fast CPU does not guarantee good commercial results - if I/O is a bottleneck, CPU will be mostly ilde. Alexei Pytel "Nathan M. Andelin" <nandelin@relational To: <MIDRANGE-L@midrange.com> -data.com> cc: Sent by: Subject: Understanding iSeries Performance midrange-l-admin@mid Positioning range.com 09/07/2001 11:31 AM Please respond to midrange-l Hi gang, I'd like to gain a better understanding of the benchmarks that IBM uses to calculate the relative CPW ratings of their boxes. I found some information by searching IBM's Web site. But it's not much. The information kind of echoed the name of the benchmark - Commercial Processing Workload. I understand the benchmarks are sort of a melding of CPU, DB I/O, and Communications, etc., rolled into one. I found that other public benchmarks offer more detail. You can even find source code for some of the programs. Some months ago, some of the list members, including me, ran a CPU bound benchmark on our machines to compare CPU speeds alone. We found, for example, that the CPU speed of a 170-2385 was about 11 times greater than a 170-2290, and the CPU speed of a 270-2250 was about 17 greater. That was interesting. For one thing, the CPW rating of the 170-2385 is about 6 times greater than the 170-2290, and the CPW rating of the 270-2250 is about 5 times greater. I had previously supposed that CPU speed and CPW rating would have been more closely correlated on the single CPU boxes. Our private benchmark quickly dispelled that erroneous assumption. Does anyone know the makeup of the benchmarks used to calculate CPW rating? Thanks, Nathan. _______________________________________________ This is the Midrange Systems Technical Discussion (MIDRANGE-L) mailing list To post a message email: MIDRANGE-L@midrange.com To subscribe, unsubscribe, or change list options, visit: http://lists.midrange.com/cgi-bin/listinfo/midrange-l or email: MIDRANGE-Lfirstname.lastname@example.org Before posting, please take a moment to review the archives at http://archive.midrange.com/midrange-l.
This mailing list archive is Copyright 1997-2014 by MIDRANGE dot COM and David Gibbs as a compilation work. Use of the archive is restricted to research of a business or technical nature. Any other uses are prohibited. Full details are available here. If you have questions about this, please contact