reiter@endor.harvard.edu (Ehud Reiter) (12/18/87)
I recently saw a videotape made by a VP at Apollo (I forget his name), who talked not only about "MIPS" and "MFLOPS", but also about "LIPS". According to a slide he put up, a "LIP" is a "Logical Inference Per Second" and is measured by the "Gabriel Benchamrk". Now, the only way I've previously heard "LIPS" used is as a measure of how many unifications per second a PROLOG system can do. The Gabriel benchmark suite, of course, measures LISP performance. Also, the Gabriel suite produces around 20-30 numbers, and Gabriel (in his book) very explicitly states that these numbers are not to be averaged or otherwise combined into one summary statistic. So, the Apollo VP (and he was a technical guy, not a marketing person) used a very misleading PROLOG term to characterize an summary statistic of LISP programs, and a summary statistic which was forbidden by the author of the benchmark suite he claimed to use. By comparison, the MIPS debate is a fountain of clarity ... Ehud Reiter reiter@harvard (ARPA,BITNET,UUCP) reiter@harvard.harvard.EDU (new ARPA)
mash@mips.UUCP (John Mashey) (12/18/87)
In article <3587@husc6.harvard.edu> reiter@harvard.UUCP (Ehud Reiter) writes: >I recently saw a videotape made by a VP at Apollo (I forget his name), who >talked not only about "MIPS" and "MFLOPS", but also about "LIPS".... .... >used a very misleading PROLOG term to characterize an summary statistic of >LISP programs, and a summary statistic which was forbidden by the author of >the benchmark suite he claimed to use. >By comparison, the MIPS debate is a fountain of clarity ... Consider how heavily LINPACK MFLOPS get used, noting that Dongarra's LINPACK report says: ``The timing information presented here should in no way be used to judge the overall performance of a computer system. The results reflect only one problem area: solving dense systems of equations using the Linpack [1] programs in a Fortran environment.'' Well, 1 out of 3 isn't bad, I guess [if that were true...sigh]. -- -john mashey DISCLAIMER: <generic disclaimer, I speak for me only, etc> UUCP: {ames,decwrl,prls,pyramid}!mips!mash OR mash@mips.com DDD: 408-991-0253 or 408-720-1700, x253 USPS: MIPS Computer Systems, 930 E. Arques, Sunnyvale, CA 94086