Path: utzoo!utgpu!news-server.csri.toronto.edu!mailrus!uwm.edu!ux1.cso.uiuc.edu!ux1.cso.uiuc.edu!aglew From: aglew@dwarfs.crhc.uiuc.edu (Andy Glew) Newsgroups: comp.arch Subject: Hot Spots Message-ID: Date: 14 Jul 90 01:32:47 GMT References: <40052@mips.mips.COM> <628@dg.dg.com> <64045@sgi.sgi.com> <1990Jul13.173331.22918@zoo.toronto.edu> <16883@dime.cs.umass.edu> Sender: usenet@ux1.cso.uiuc.edu (News) Organization: University of Illinois, Computer Systems Group Lines: 10 In-Reply-To: yodaiken@chelm.cs.umass.edu's message of 14 Jul 90 00:15:58 GMT Or how about that other fallacy related to the hot-spot principle? "It isn't worth making any performance change, unless that single change, by itself, will improve the performance of the system by (you choose your value) 10%, 30%, 50%, 100%..." A large number of small, easy to make, performance improvements can easily add up to a large performance improvement. -- Andy Glew, aglew@uiuc.edu