Path: utzoo!yunexus!geac!syntron!jtsv16!uunet!lll-winken!lll-tis!ames!husc6!hscfvax!pavlov
From: pavlov@hscfvax.harvard.edu (G.Pavlov)
Newsgroups: comp.arch
Subject: Re: MIPS Performance Brief 3.5, October 1988 [more than long]
Keywords: benchmarks
Message-ID: <648@hscfvax.harvard.edu>
Date: 28 Oct 88 05:45:09 GMT
Article-I.D.: hscfvax.648
References: <7000@winchester.mips.COM>
Organization: Health Sciences Computing Facility, Harvard University
Lines: 25

In article <7000@winchester.mips.COM>, mash@mips.COM (John Mashey) writes:
> People have been beating me up for this, so here it is.  As usual,
> let me know if you have better numbers for things.  We're always trying

  I learned some (to me) interesting things from this article; thank you.

  An area of performance, admittedly complex, but one which has increasing-
  ly interested me, is how well a processor holds up in a multitasking en-
  vironment.  For instance, I have just completed comparing the performance
  of a "well-known" machine (NOT MIPS), that is sold both as a workstation
  and a server, against a Microvax II.  The tasks were all Ingres jobs, so
  the executables were large.

  When one to several tasks were executed, this machine showed a performance
  ratio of 6-8:1 to the uVax II.  Beyond that, it slid almost immediately to
  approx. a 3:1 ratio.  This is consistent with what I have seen published
  about this machine, so I wasn't surprised by the results.

  But this type of result tends to color my view of single-process benchmarks,
  no matter how well conceived or carefully executed.  I realize that large
  numbers of simultaneous "jobs" are not typical in a workstation environment,
  but then again multitasking is common. 

  Has there been any work addressing this ?

  greg pavlov, fstrf, amherst, ny