NetNews Usenet Archive 1992 #16/NN_1992_16.iso/spool/comp/unix/aix/8085

Internet Message Format | 1992-07-22 | 1.9 KB

Path: sparky!uunet!zaphod.mps.ohio-state.edu!cs.utexas.edu!ut-emx!ibmchs!auschs!awdprime.austin.ibm.com!chukran.austin.ibm.com!rudy
From: [e-mail]
Newsgroups: comp.unix.aix
Subject: Re: need help with SPEC 008.espresso benchmark on IBMR2-560F
Keywords: SPEC, 560F
Message-ID: <[e-mail]>
Date: 21 Jul 92 17:36:14 GMT
References: <[e-mail]>
Sender: [e-mail] (USENET News)
Reply-To: [e-mail](Rudy Chukran)
Organization: IBM Advanced Workstation Division
Lines: 20

In article <[e-mail]>, [e-mail] (Vic Abell) writes:
|> Has anyone run the SPEC 1.0a or 1.2b benchmark 008.espresso on an R2 560F
|> and gotten a number as small as the one IBM reports in the SPEC Newsletter?
|> (I know, I know, I should be running C{INT,FP}92, but my March order has
|> yet to produce a tape by virtue of some bumbling here.)  The best elapsed
|> time I get is 64.0 seconds, while IBM reports 54.3.  Since I can reproduce
|> the other nine results, I don't think some time gobbling process is lurking
|> in the background and snickering while it inflates my run times.  I use
|> the "-O" optimization that IBM reports in the March, 1992 SPEC Newsletter.

This is a wild guess:  Did you try linking these statically?  The AIX FAQ
tells how.   Static linking sometimes speeds up cpu bound processes.
If espresso uses libc or libm, or other libs a lot, this might help.
*********************************************************************
Rudy Chukran                |       EMAIL:
IBM AIX Technical Consulting| RSCS: CHUKRAN at AUSTIN
11400 Burnet Rd.            | AWDnet:[e-mail]
Internal ZIP 2830           | internet: [e-mail]
Austin, Texas 78758         |
*********************************************************************