Return to BSD News archive
Path: euryale.cc.adfa.oz.au!newshost.carno.net.au!harbinger.cc.monash.edu.au!news.cs.su.oz.au!inferno.mpx.com.au!nsw1.news.telstra.net!wa.news.telstra.net!news.telstra.net!news-out.internetmci.com!newsfeed.internetmci.com!news.mathworks.com!enews.sgi.com!fido.asd.sgi.com!refugee.engr.sgi.com!sca From: sca@refugee.engr.sgi.com (Steve Alexander) Newsgroups: comp.unix.bsd.freebsd.misc,comp.unix.bsd.bsdi.misc,comp.sys.sgi.misc Subject: Re: no such thing as a "general user community" Date: 3 Apr 1997 02:14:26 GMT Organization: Silicon Graphics, Inc., Mountain View, CA Lines: 28 Message-ID: <5hv3q2$f2k@fido.asd.sgi.com> References: <331BB7DD.28EC@net5.net> <5hun4a$eav@fido.asd.sgi.com> <5hus36$3ev$1@bofh.noc.best.net> <5huvun$hlq@fido.asd.sgi.com> Reply-To: sca@sgi.com NNTP-Posting-Host: fddi-refugee.engr.sgi.com Xref: euryale.cc.adfa.oz.au comp.unix.bsd.freebsd.misc:38353 comp.unix.bsd.bsdi.misc:6570 comp.sys.sgi.misc:29643 In article <5huvun$hlq@fido.asd.sgi.com>, Larry McVoy <lm@slovax.engr.sgi.com> wrote: >I beg to differ. LADDIS, TPC, Webstone, lmbench, etc., are all >benchmarking combinations of hardware and operating systems. In many >cases, it is impossible to seperate the two, there is only one OS for >some platforms. It's actually worse than that. Some benchmarks such as TPC and SPECweb don't even necessarily have the same networking configuration, the same load generators, or the same user-level application. Everything can be different from one vendor's results to another's; if you're looking for "only change one variable like my sixth-grade science teacher said," forget it. Those benchmarks are effectively designed to produce a price/performance number. "This configuration can produce some number of ops/sec. It costs $X." TPC actually seems to report $/TPC; they must be more honest ;->. Customers seem to use this information to try to figure out how to get maximized performance for minimized investment. They're useful benchmarks for that (assuming that you think that they measure real-world workloads; that's a different issue). They also try to ensure reproducibility by requiring vendors to report enough information that a customer could generate the same results given the same setup. -- Steve Alexander | Silicon Graphics, Inc. | +1 (415) 933-6172 (Voice) sca@sgi.com | http://reality.sgi.com/sca | +1 (415) 933-0513 (FAX) "I'm going outside and I'm gonna learn Perl."