Path: utzoo!utgpu!news-server.csri.toronto.edu!cs.utexas.edu!usc!jarthur!uunet!stanford.edu!leland.Stanford.EDU!elaine18.Stanford.EDU!dhinds
From: dhinds@elaine18.Stanford.EDU (David Hinds)
Newsgroups: comp.sys.sgi
Subject: Re: Weird Fortran i/o?
Message-ID: <1991May3.225537.7216@leland.Stanford.EDU>
Date: 3 May 91 22:55:37 GMT
References: <9105031951.AA02478@smaug.physics.mcgill.ca>
Sender: news@leland.Stanford.EDU (Mr News)
Organization: Stanford University - AIR
Lines: 23

In article <9105031951.AA02478@smaug.physics.mcgill.ca> igraham@SMAUG.PHYSICS.MCGILL.CA (Ian Graham) writes:
>
>A user on our system was checking benchmarks for system i/o on 
>IRIS/SUN/HP9000/DECSTATION and came up with the following quirky 
>results: (THe code is given at the bottom)  THe program writes
>100000 bunches of 4 integers into a direct access fie, rewinds, 
>and reads 'em all back.   Here are the results for a 4D-25:
>
>START-TIME         0.0100      0.0700
>IO WRITE           8.9400      4.7600
>IO READ           26.4600    353.0300 <<---!!!!!!
>
>So what gives?  Why should it take 5 MINUTES to read some data written
>in 5 SECONDS!

    I don't know why reads in particular are so slow, but my experience
was that ALL fortran I/O was really slow.  I didn't make detailed timings,
but I remember that one program which took about an hour to read a large
unformatted file ran in about 3 minutes when I replaced a single Fortan
READ statement with a C function to read a short string of bytes.

 -David Hinds
  dhinds@cb-iris.stanford.edu