Path: utzoo!utgpu!news-server.csri.toronto.edu!rpi!uupsi!njin!gould From: gould@pilot.njin.net (Brian Jay Gould) Newsgroups: comp.arch Subject: Re: Novice question: measuring speed Summary: The only way to measure speed Message-ID: Date: 19 Mar 91 02:39:28 GMT References: <645@ssdc?> Organization: NJ InterCampus Network, New Brunswick, N.J. Lines: 22 There was a question a few years ago about what is the fastest computer... Fortunately, we were able to get away with the explanation that all system performance is very application dependent. In fact, some of the best CPU performers are so poorly balanced in I/O and other features that they benchmark behind slower CPU's. The same applies to supercomputer architectures where a system mey work well with very large vectors but has such a high latency that most scalar applications show up as slower than a PC. The only real way to measure the speed of a system is to drop it from a building and calculate 42 feet per second squared. -- -- Any disclaimers made for me, by me, or about me - may or may not accurately reflect my failure to be reflecting the opinions of myself or anyone else. ************************************************* * Brian Jay Gould - Professional Brain-stormer * *************************************************