home *** CD-ROM | disk | FTP | other *** search
- Newsgroups: comp.arch
- Path: sparky!uunet!ferkel.ucsb.edu!taco!gatech!swrinde!elroy.jpl.nasa.gov!decwrl!pa.dec.com!nntpd2.cxo.dec.com!nntpd.lkg.dec.com!ryn.mro4.dec.com!wrksys.enet.dec.com!bhandarkar
- From: bhandarkar@wrksys.enet.dec.com (Dileep Bhandarkar)
- Subject: Re: MIPS R4400
- Message-ID: <1992Nov9.210442.5149@ryn.mro4.dec.com>
- Sender: news@ryn.mro4.dec.com (USENET News System)
- Organization: Digital Equipment Corporation
- References: <lflj8pINN158@spim.mti.sgi.com> <lftam5INNgoi@exodus.Eng.Sun.COM>
- Date: 9 NOV 92 16:04:06
- Lines: 19
-
-
- In article <lftam5INNgoi@exodus.Eng.Sun.COM>, walter@zarquon.Eng.Sun.COM (Walter Bays) writes...
- >>In article <JCALLEN.92Nov6135200@marley.think.com> jcallen@marley.think.com (Jerry Callen) writes:
- >>[I was surprised to see MIPS use SPEC 89 numbers...]
- >
- >I also wonder why you would use the old 89 suite. Was it that some of
- >the new 92 benchmarks are too complicated to simulate, in particular
- >026.compress and 072.sc? (I believe there are ways around the
- >difficulties.) But even if SPECint92 was intractable, can't you at
- >least simulate SPECfp92?
- >
- Simulation is slow and often difficult when a benchmark does I/O. It is a lot
- easier to do 10 benchmarks that have been around for 3 years, than it is to do
- 20 relatively new ones. With SPEC89 numbers in detail, one can make reasonable
- guesses at SPECint92 and SPECfp92.
-
- While more would be better, it is nice to see some numbers with backup data.
-
- Dileep
-