home *** CD-ROM | disk | FTP | other *** search
- Xref: sparky comp.arch:10444 comp.lang.misc:3516
- Newsgroups: comp.arch,comp.lang.misc
- Path: sparky!uunet!charon.amdahl.com!pacbell.com!decwrl!sdd.hp.com!cs.utexas.edu!sun-barr!ames!saimiri.primate.wisc.edu!usenet.coe.montana.edu!giac1.oscs.montana.edu!osycs
- From: osycs@giac1.oscs.montana.edu (Craig Spannring)
- Subject: Re: Hardware Support for Numeric Algorithms
- Message-ID: <1992Nov6.075459.13348@coe.montana.edu>
- Keywords: n
- Sender: usenet@coe.montana.edu (USENET News System)
- Organization: Geographic Information & Analysis Center Montana State University
- References: <1992Oct23.004313.29196@ntuix.ntu.ac.sg> <1992Oct29.153514.22927@yrloc.ipsa.reuter.COM> <1992Nov5.202412.7266@linus.mitre.org>
- Date: Fri, 6 Nov 1992 07:54:59 GMT
- Lines: 30
-
-
- In article <1992Nov5.202412.7266@linus.mitre.org> bs@gauss.mitre.org (Robert D. Silverman) writes:
- >You've probably never done any large scale computing.
- >
- >I measure the run time of my algorithms in MIPS-YEARS. A MIPS-YEAR
- >is a 1 MIPS machine running for 1 year, or approximately 3.1 x 10^13
- >instructions. Jobs that take several hundred MIPS-YEARS are not uncommon.
-
- Large scale computing demands that you to avoid the unreadable
- micro-optimizations that Herman is suggesting. If one of your little
- attempts at optimizing the code ends up producing incorrect results
- then you've just wasted months of research time. Meanwhile your
- colleagues that have programs producing correct results are getting
- published and getting grants. How much research and computing time are
- you wasting with buggy programs?
-
- The second problem with these 'optimizations' is that they aren't.
- Herman seems to be most concerned with saving a few clock cycles per
- iteration. The only significant performance gains will be found be
- using an algorithm that uses fewer iterations. Example- Even a highly
- optimized bubble sort will run slower than an unoptimized quick sort
- for the large data sets you are refering to.
-
- Get a clue.
-
- --
- =========================================================================
- Six of one, 110 (base 2) of | Craig Spannring --- (406) 994-6128
- another. | osycs.giac1.oscs.montana.edu
- ----------------------------+ Geographic Information & Analysis Center
-