evans@mhuxt.UUCP (05/29/87)
Yesterday I had a chance to compare a Mac II (it was at my local dealer's shop for almost 3 hrs before its new owner came!) and my Prodigy Prime (on my Mac+). I realize this isn't much of a comaprison, but it is interesting and I don't have a 68020/68881 option on my C compiler (yet). I used Steven Eubank's mandelbrot set (which was compiled to take advantage of the 68881 and 68020). The Prodigy Prime was almost 2x as fast as the Mac II!!! The II was in the simple (2 grey level) display mode and the program doesn't access the disk. Can anyone explain the difference? (I can understand a 10 or 20% difference, but 100% is hard to swallow). The guy at the store loaded up a Mac spin demo and it was qualitatively slower than on the P' at home - Steve Crandall ihnp4!mhuxt!evans
olson@harvard.UUCP (Eric Olson) (06/01/87)
In article <1740@mhuxt.UUCP> evans@mhuxt.UUCP (Steve Crandall) writes: >The Prodigy Prime was almost 2x as fast as the Mac II!!! The II was in the >simple (2 grey level) display mode and the program doesn't access the disk. >Can anyone explain the difference? (I can understand a 10 or 20% difference, >but 100% is hard to swallow). Part of the time difference is that the Prodigy, when properly configured (i.e., use Mac RAM for RAMDisk, Prodigy RAM for RAM), has zero wait state memory. The Mac II, I believe, has one wait state memory. Also, if the tests are graphically oriented, keep in mind that the Mac II has a larger screen (if a full-screen graphics test) and Color quickdraw bottlenecks in various places. -Eric
ali@rocky.STANFORD.EDU (Ali Ozer) (06/01/87)
In article <1090@harvard.UUCP> olson@harvard.UUCP (Eric Olson) writes: >In article <1740@mhuxt.UUCP> evans@mhuxt.UUCP (Steve Crandall) writes: >>The Prodigy Prime was almost 2x as fast as the Mac II!!! >>Can anyone explain the difference? (I can understand a 10 or 20% difference, >>but 100% is hard to swallow). >Part of the time difference is that the Prodigy, when properly configured >(i.e., use Mac RAM for RAMDisk, Prodigy RAM for RAM), has zero wait state >memory. ... Also the Mac II has a larger >screen (if a full-screen graphics test) and Color quickdraw bottlenecks in >various places. According to some values passed on to me by a friend, the savage benchmark runs considerably (> 2x) faster on the 14 MHz Turbo Amiga than on the 16 MHz Mac II. (This is taking into consideration the effects of the languages used by looking at the performance of the same program on the Mac Plus vs the 68000 Amiga.) The reason must lie in the fact that the 68020 in the Mac II has to do a lot more housekeeping work than the 68020 in the Prodigy or the Turbo Amiga. (In the case of the Turbo Amiga all graphics operations are still done by the 68000/blitter combo and the 68020 is free to do whatever it wants in its 32 bit memory at the same time.) [ps. I might be wrong here --- Maybe it has to do with the fact that the language used for the Mac II was not as good when it came to generating code for the 68020/68881 combo. Just for reference, the savage benchmark value quoted to me for the Mac II was 4 seconds.] Ali Ozer, ali@rocky.stanford.edu