Actually, there are two ways to measure a flop: - A mult _or_ an add - A mult _and_ an add I believe that the "old" way is the latter and the "new" way is the former, though I'd have to check my Numerical Analysis book to make sure. Intel's chips aren't really capable of many flops in comparison to some of the other beastly CPUs out there (e.g. DEC (compaq) Alpha - the king of flops, IBM's Power series cpus, SGI's R10k, etc) Gabe On Wed, Jul 12, 2000 at 03:23:06PM -0500, Luke Francl wrote: > On Wed, 12 Jul 2000, Timothy Wilson wrote: > > > On Wed, 12 Jul 2000, Philip C Mendelsohn wrote: > > > > > It's on the order of millions -- hence Mflops > > > > I wasn't sure that 1 flop == 1 multiplication/sec > > It isn't, necessarily. A flop is a floating point operation (anything with > decimals in it). I don't know if it includes addition/subtraction or just > mutliplication/division. > > Are you looking for pure integer performance? > > > > > --------------------------------------------------------------------- > To unsubscribe, e-mail: tclug-list-unsubscribe at mn-linux.org > For additional commands, e-mail: tclug-list-help at mn-linux.org -- -------------------------------------------------------------------------------- Gabe Turner X-President, ACM @ U of MN dopp at acm.cs.umn.edu "Be nice to him [Kowalski] because he's the product of your loins!" -- Stimpy "My fake loins!!" -- Ren Hoek - Ren takes on fatherhood in "Fake Dad" -------------------------------------------------------------------------------- --------------------------------------------------------------------- To unsubscribe, e-mail: tclug-list-unsubscribe at mn-linux.org For additional commands, e-mail: tclug-list-help at mn-linux.org