Xref: utzoo comp.arch:14512 comp.sources.wanted:10878 Path: utzoo!utgpu!jarvis.csri.toronto.edu!cs.utexas.edu!uunet!indetech!fiver!palowoda From: palowoda@fiver.UUCP (Bob Palowoda) Newsgroups: comp.arch,comp.sources.wanted Subject: Re: Byte UNIX Benchmarks Message-ID: <1020@fiver.UUCP> Date: 8 Mar 90 19:18:14 GMT References: <22943@mimsy.umd.edu> Distribution: comp Organization: Fiver Communications Fremont, Ca Lines: 38 From article <22943@mimsy.umd.edu>, by mjr@atreus.umiacs.umd.edu (Marcus J. Ranum): > In article <1990Mar6.163657.7550@me.toronto.edu> yap@me.utoronto.ca (Davin Yap) writes: >>Has anyone got a copy of the UNIX benchmarks discussed in the latest >>issue of Byte magazine. > > are these benchmarks from the magazine that ues > Sieve of Eratosthenes as a compiler benchmark, and a Loutus-1-2-3 > spreadsheet as a CPU benchmark ? > > Or is this the infamous Jerry Pournelle BASIC benchmark re-written > in C? > Neither, It seems to be a complete rewrite in C. It takes into account UNIX's multiple processes when bench testing. Here's some of the tests: Dhrystone 2 (with and without register vars) Sequential and Random memory access with incremental array sizes Arithmatic tests (short, long, float, doubles, ints) Process tests (system call overhead, pipe throughput, pipe switch process creation, execl throughput) File system throughput (incremental sizes) Client/Server Database throughput (incremental number of client/servers) I did post the sources, along with compiled executeables for Xenix386 and UNIX 386 on my bbs. The file is 'Bbench.tar.Z' and is located in the unix_bench files area of the bbs. It is approx 800K, so please don't ask me to mail it. ---Bob P.S. -- Bob Palowoda indetech!fiver!palowoda *Home of Fiver BBS* login: bbs Home {sun|daisy}!ys2!fiver!palowoda (415)-623-8809 1200/2400 Work {sun|pyramid|decwrl}!megatest!palowoda (415)-623-8806 2400/9600/19200 TB Voice: (415)-623-7495 palowoda@fiver Public access UNIX XBBS