Path: utzoo!utgpu!news-server.csri.toronto.edu!cs.utexas.edu!samsung!zaphod.mps.ohio-state.edu!usc!apple!agate!darkstar!maths.tcd.ie
From: ajudge@maths.tcd.ie (Alan Judge)
Newsgroups: comp.os.research
Subject: Re: Toy Experiment with Shared Memory Programming
Message-ID: <8915@darkstar.ucsc.edu>
Date: 13 Nov 90 09:58:27 GMT
Sender: usenet@darkstar.ucsc.edu
Organization: Dept. of Maths, Trinity College, Dublin, Ireland.
Lines: 29
Approved: comp-os-research@jupiter.ucsc.edu

In <8754@darkstar.ucsc.edu> jms@central.cis.upenn.edu (Jonathan M. Smith) writes:
>Some Notes on Experiences programming shared memory

Just out of interest, I thought that I would try piper and mem on a
different machine type.  It should give a better idea of the effects of
differing kernel implementations on the performance.

Below are the results of piper and mem run on a DECsystem 5810 (MIPS
based) running Ultrix 3.1C.  The system has 32Mb of memory and is
running from local disks.  The size of the buffer cache should remove
the effects of the disks after the first execution.  /vmunix, the test
file used, is 2.8Mb in size.

parade 1% time pipe < /vmunix > /dev/null
0.0u 2.0s 0:05 39% 3+4k 1+0io 0pf+0w av:8k m:4k refpf: 0
parade 2% time pipe < /vmunix > /dev/null
0.0u 2.0s 0:02 67% 3+4k 0+0io 0pf+0w av:8k m:4k refpf: 0
parade 2% time mem < /vmunix > /dev/null
0.0u 1.1s 0:01 80% 3+2k 3+1io 0pf+0w av:6k m:5k refpf: 0
parade 3% time mem < /vmunix > /dev/null
0.0u 1.1s 0:01 91% 3+2k 0+0io 0pf+0w av:6k m:5k refpf: 0

It is interesting to note that, in this system, shared memory is
almost twice as fast as pipes.
-- 
Alan Judge   ajudge@maths.tcd.ie  a.k.a. amjudge@cs.tcd.ie +353-1-772941 x1782

File names are infinite in length where infinity is set to 255 characters.
                                  - Peter Collinson, "The Unix File System"