home *** CD-ROM | disk | FTP | other *** search
- Newsgroups: comp.benchmarks
- Path: sparky!uunet!stanford.edu!ames!data.nas.nasa.gov!wilbur!eugene
- From: eugene@wilbur.nas.nasa.gov (Eugene N. Miya)
- Subject: Re: SPECmarks and Multiprocessor Machines
- References: <1992Aug11.195901.10432@crd.ge.com> <BsuAp1.KLK@cup.hp.com> <l8gph3INNb7b@exodus.Eng.Sun.COM>
- Sender: news@nas.nasa.gov (News Administrator)
- Organization: NAS, NASA Ames Research Center, Moffett Field, CA
- Date: Wed, 12 Aug 92 17:27:35 GMT
- Message-ID: <1992Aug12.172735.3318@nas.nasa.gov>
- Lines: 14
-
- It should be pointed out that many of the original SPEC suite programs,
- those floating point intensive things, were numeric codes which were
- vectorizable (a form of parallelism). That's not a clear cut advantage,
- I know of some disadvantages, but you should be aware that parallelism
- was inherent in the suite long before now. The trick is in understanding
- what parallelism can AND CAN'T do for you.
-
- Go look for those non-linearities.
-
- --eugene miya, NASA Ames Research Center, eugene@orville.nas.nasa.gov
- Resident Cynic, Rock of Ages Home for Retired Hackers
- {uunet,mailrus,other gateways}!ames!eugene
- Second Favorite email message: Returned mail: Cannot send message for 3 days
- A Ref: Mathematics and Plausible Reasoning, vol. 1, G. Polya
-