LIALG@CUNYVM.CUNY.EDU (11/21/88)
DEAR READER, i WOULD LIKE INFORMATION ON THE DEFINITION OF LIPS (I.E., LOGICAL INFERENCES PER SECOND) AND ANY BENCHMARKS THAT ARE 'STANDARD' IN MEASURING IT. THANK YOU, LARRY. #! rnews
ok@quintus.uucp (Richard A. O'Keefe) (11/23/88)
In article <1732LIALG@CUNYVM> LIALG@CUNYVM.CUNY.EDU writes: > i WOULD LIKE INFORMATION ON THE DEFINITION OF LIPS (I.E., LOGICAL > INFERENCES PER SECOND) AND ANY BENCHMARKS THAT ARE 'STANDARD' IN > MEASURING IT. The only official definition of LI/s is let T be the time in seconds to run naive reverse of a 30-element list N times, for N large enough that timing fluctuations are small. The LI/s figure is then T/(496*N). *NO* other benchmark measures LI/s; anything else is "procedure calls per second for benchmark X". Naive reverse is of course nrev([], []). nrev([X|Xs], R) :- nrev(Xs, R1), append(R1, [X], R). append([], L, L). append([H|T], L, [H|R]) :- append(T, L, R).
cdsm@ivax.doc.ic.ac.uk (Chris Moss) (11/28/88)
In article <721@quintus.UUCP> ok@quintus.UUCP (Richard A. O'Keefe) writes: >In article <1732LIALG@CUNYVM> LIALG@CUNYVM.CUNY.EDU writes: >> i WOULD LIKE INFORMATION ON THE DEFINITION OF LIPS (I.E., LOGICAL >> INFERENCES PER SECOND) AND ANY BENCHMARKS THAT ARE 'STANDARD' IN >> MEASURING IT. > >The only official definition of LI/s is ... This is an interesting non-answer to the stated question. Is the "official" version "LI/s" or "LIPS"? I am unable to find the original definition which was I believe made in a paper by Furukawa from ICOT in around 1982. We don't seem to have a copy here and an enquiry to ICOT produced no answer. Does anyone have that paper and if so could they clarify what the name he gave it? Just an idle curiosity! Chris Moss, Imperial College
ok@quintus.uucp (Richard A. O'Keefe) (11/30/88)
In article <510@gould.doc.ic.ac.uk> cdsm@doc.ic.ac.uk (Chris Moss) writes: >In article <721@quintus.UUCP> ok@quintus.UUCP (Richard A. O'Keefe) writes: >>In article <1732LIALG@CUNYVM> LIALG@CUNYVM.CUNY.EDU writes: >>> i WOULD LIKE INFORMATION ON THE DEFINITION OF LIPS (I.E., LOGICAL >>The only official definition of LI/s is >... > >This is an interesting non-answer to the stated question. >Is the "official" version "LI/s" or "LIPS"? It is certainly interesting to find an answer that included source code for the benchmark in question described as a "non-answer". LI/s is my personal spelling of logical inferences / second, and I'm proud of it! (I also pronounce it "ceidlidhs", but that's a joke.) The published use of this particular benchmark dates back to 1977.
cdsm@ivax.doc.ic.ac.uk (Chris Moss) (12/05/88)
>>This is an interesting non-answer to the stated question. >>Is the "official" version "LI/s" or "LIPS"? > >It is certainly interesting to find an answer that included source code >for the benchmark in question described as a "non-answer". > I was referring to the use of the name, not the benchmark! >LI/s is my personal spelling of logical inferences / second, >and I'm proud of it! (I also pronounce it "ceidlidhs", but that's a joke.) > >The published use of this particular benchmark dates back to 1977. Yes, I know the use of the naive reverse benchmark first appeared in Dave Warren's paper on the Edinburgh compiler. But it was not distinguished there from a large number of other tests nor was it given a name. I was interested in the source of the name "LIPS" which seems to be used by everyone (except Richard). Can anyone help? Chris Moss.