From the Lisp-FAQ:
> Gabriel Lisp Benchmarks are available by anonymous ftp as
> ai.toronto.edu:/pub/gabriel-lisp-benchmarks.tar.Z.
> The benchmarks are described in the book "Performance Evaluation of
> Lisp Systems", by Richard Gabriel.
I would highly suggest not to use the gabriel benchmarks to compare
machines. As I understand, they are intended to be useful for
optimizing compilers. They test specific aspects of isolated features
of a Lisp system. I found the results to be very misleading and not
useful to determine how fast any real-world application (of mine)
would be on a tested platform.
Even comparing different compilers on the same machine didn't look
much better.
BTW, I would really like to see a set of standard benchmarks in the
form of 5 to 8 'real' applications so one could pick those which
reflect his needs best. As it seems, there is interest in this
newsgroup for such data.
Suppose I collect some applications, are there enough people who would
run them on their systems? Total run time would be several hours, and
maybe there is some work to do for specific platforms, say half an
hour. If you would do that, please drop me a note with a description
of the systems you would run the benchmarks on.
If there is a useful base of promised data, I make a 'Call for
applications'. I fear there are not that much applications that are
free, optimized for high-speed, but portable and easy to run, but
we'll see then.
--
%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
Martin Cracauer <········@wavehh.hanse.de>,Voice+4940-5221829,Fax.-5228536
Waldstrasse 200, 22846 Norderstedt, Germany. German language accepted.
;;; Happily using arms, legs and sometimes a brain