Path: utzoo!utgpu!news-server.csri.toronto.edu!cs.utexas.edu!wuarchive!uunet!unisoft!fai!sequent!dafuller
From: dafuller@sequent.UUCP (David Fuller)
Newsgroups: comp.benchmarks
Subject: Re: Advice on benchmark sought
Keywords: Dhrystones
Message-ID: <51032@sequent.UUCP>
Date: 20 Jan 91 06:31:39 GMT
References: <1991Jan16.210617.6484@agate.berkeley.edu>
Reply-To: dafuller@sequent.UUCP (David Fuller)
Distribution: comp
Organization: Sequent Computer Systems, Inc
Lines: 35

In article <1991Jan16.210617.6484@agate.berkeley.edu> ilan343@violet.berkeley.edu (Geraldo Veiga) writes:
>I am trying to assemble a set of publicly available benchmarks to run
>on UNIX/386 machines.  I am starting small.  I got copies of Dhrystone,
>Whetstones and Linpack from NETLIB.
>
>Here is the problem:
>
>Running dhrystone 2.1 (c version) on a 486 machine running ISC 2.2, I
>get all kinds of crazy results, with Dhrystone ratings ranging from
>3000 to 16000 in different runs.  The same machine under DOS, running
>the Drystones version in QAPLUS rates (consistently) at 15172.
>
>Question:
>
>Is this normal?  Is the hardware or software flaky?  If I time the run
>with Sysv's "time", user time also show the same wide variation.

You have given no indication that your initial conditions have been analyzed
and corrected for possible external influence.

While significant single-run differences can be attributed to sterling
initial conditions and a bad cache line arrangements, your results are 
worth breaking down to the source and completely reconsidering.

>
>
>Suggestions appreciated.

Examine your initial conditions.  Eliminate every variable you can and
document those you can't mess with.
-- 
Dave Fuller				   
Sequent Computer Systems		  Think of this as the hyper-signature.
(708) 318-0050 (humans)			  It means all things to all people.
dafuller@sequent.com