NU123952@VM1.NoDak.EDU (Mark A. Ordal) (02/06/91)
This is the first time I've heard allegations that SPEED or FAST can lead to invalid numerical results. In my own tests I have NEVER found the numerical results to be different. Incidentally, the program I benchmarked had the following run times: HP28C 24 seconds HP28S 16 seconds HP48SX 11.5 seconds HP28S (FAST) 8 seconds The numerical results were identical in all 4 cases. Dr. Mark A. Ordal Physics Department North Dakota State University Fargo, ND 58105 NU123952@NDSUVM1
edp@jareth.enet.dec.com (Eric Postpischil (Always mount a scratch monkey.)) (02/08/91)
In article <09E78DB1800015A0@gacvx2.gac.edu>, NU123952@VM1.NoDak.EDU (Mark A. Ordal) writes: >This is the first time I've heard allegations that SPEED or FAST can lead >to invalid numerical results. In my own tests I have NEVER found the >numerical results to be different. Test it with the batteries heading toward depletion, but not yet tripping the low-battery indication. -- edp (Eric Postpischil) "Always mount a scratch monkey." edp@jareth.enet.dec.com