kirchner@uklirb.UUCP (Reinhard Kirchner) (11/21/88)
From article <7722@aw.sei.cmu.edu>, by firth@sei.cmu.edu (Robert Firth): > > Too bad he didn't base his decision in part on whether users performing > floating-point calculations prefer better answers to worse ones. Too > bad also that the 'megaflops per millisecond' benchmarking culture tends > to be heavily biased in favour of "wrong answers fast". This is what we say since many years: A CRAY is a machine to produce wrong results very fast. But not the CRAY only, others do also, but not so FAST! To check the results of Your CRAY, CYBER, VPxxx, etc. You may use Fortran-SC or Pascal-SC -:) ( not as fast because nobody found accurate results worth building the neccessary hardware but proven accurate! ) I can post details of this on request. R. Kirchner