OKeefe.R.A.%EDXA@sri-unix.UUCP (11/05/83)
From: O'Keefe HPS (on ERCC DEC-10) <OKeefe.R.A. at EDXA>
Someone recently said in this Digest that Waterloo Prolog was
an interpreter that was as fast as DEC-10 Prolog. Now I have heard
nothing but good about Waterloo Prolog, and have recommended it to
a couple of people, so don't take this as any sort of attack on it.
The thing is, you've got to take the machine into consideration
as well. Here are some figures to show what's happening:
Dec-10 compiled on KI-10/Bottoms-10 20k LIPS
C-Prolog v1.4.EdAI on VAX 750/UNIX 1k LIPS
C-Prolog v1.4.EdAI on VAX 780/VMS 1.8kLIPS (estimate)
C-Prolog (??version??) IBM 3081/(??) 11k LIPS
If Waterloo Prolog is absolutely as fast as DEC-10 Prolog,
then it is relatively about twice as fast as C-Prolog. Given
that C-Prolog is written in C, and has not been tuned for IBMs,
and that Waterloo Prolog is said to be written in assembler,
this is a good but not earth-shaking result. A compiler of the
same quality might be expected to yield 200k LIPS.
What we need is more systems that are at least twice as
fast as DEC-10 Prolog. People without 3081s are still waiting
for one that is half as fast. (If yours is that good or better,
PLEASE tell us.)