Date: Tue, 30 Apr 1996 09:16:17 -0700 (PDT) From: "Jonathan M. Bresler" <jmb> To: msmith@atrad.adelaide.edu.au (Michael Smith) Cc: andrew@sedona.net, freebsd-questions@freebsd.org Subject: Re: tests Message-ID: <199604301616.JAA18364@freefall.freebsd.org> In-Reply-To: <199604300143.LAA18051@genesis.atrad.adelaide.edu.au> from "Michael Smith" at Apr 30, 96 11:13:34 am
next in thread | previous in thread | raw e-mail | index | archive | help
Michael Smith wrote: > > Andrew S. Clapp stands accused of saying: > > > > Anybody know where to find a benchmark package like dhry? > > We are pitting solaris-x86 against freebsd! :-) Oughtta > > be a interesting race as sunsoft says that 2.5 is "supposed" > > to be competitive. > > Why would an operating system significantly bias a CPU speed benchmark? most CPU benchmarks are not CPU benchmarks--they are CPU and memory subsystem benchmarks. the OS affects the benchmarks access to memory. a benchmark tha shows this beautiully is Hint from ameslab. the same taks is run repeatedly, each time with a larger set size. the results are a graph of speed vs memory footprint. most machines showhte size of on-chip cache, L2 cahce, memory size as steps in the graph. each step lower than the previous one as the problem size access slower and slower resources. http://www.scl.ameslab.gov/scl/HINT/HINT.html jmb -- Jonathan M. Bresler FreeBSD Postmaster jmb@FreeBSD.ORG FreeBSD--4.4BSD Unix for PC clones, source included. http://www.freebsd.org/
Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?199604301616.JAA18364>