piet@ruuinf (Piet van Oostrum) (12/17/88)
Today I tried to run a Pascal comparison Sun 3/60 vs HP9000/318. The results are very surprising: The Sun3/60 is a 20 MHz 68020, The HP9000/318 is a 16.7 MHz 68020. The speed diffrence of compilation and C programs is 20-50% in favor of the Sun. So far so good. But: The speed of a translated Pascal program on the Sun is 6-7 times the speed of the program on HP. (I used the TeX tangle program as an example). Both are compiled without optimizations and with runtime checks (which doesn't make much difference). I can't check the figures for HP, because I don't have one at hand, so I must trust what I wrote down, but tangling tangle.web tangle.ch took about 34 seconds on the HP and 5.2 seconds on Sun. WHY IS THE HP PASCAL CODE SO SLOW???? Does anybody have experience with this? -- Piet van Oostrum, Dept of Computer Science, University of Utrecht Padualaan 14, P.O. Box 80.089, 3508 TB Utrecht, The Netherlands Telephone: +31-30-531806 UUCP: ...!mcvax!hp4nl!ruuinf!piet
irf@kuling.UUCP (Bo Thide) (01/02/89)
In article <921@ruuinf.UUCP> piet@ruuinf (Piet van Oostrum) writes: > >WHY IS THE HP PASCAL CODE SO SLOW???? > >Does anybody have experience with this? What HP-UX version did you use? How much memory did you have in your Sun? (The 318 - now replaced by the much faster 340 - has only 4 MB RAM.) I don't have Pascal myself but have tested Sun 3/2x0 C and FORTRAN77 versus HP9000/350. C programs seem to execute more or less at the same speed whereas FORTRAN programs on the HP are up to 3-4 times *faster*! So your results are very odd/interesting... -Bo