Path: utzoo!utgpu!news-server.csri.toronto.edu!cs.utexas.edu!wuarchive!psuvax1!rutgers!aramis.rutgers.edu!remus.rutgers.edu!declan From: declan@remus.rutgers.edu (Declan McCullagh/LZ) Newsgroups: comp.sys.mac.misc Subject: Re: The New Macs: Greedy Compromises? Message-ID: Date: 9 Dec 90 22:38:23 GMT References: <1990Dec5.072931.4079@cs.ucla.edu> <11447@goofy.Apple.COM> <47215@apple.Apple.COM> Organization: Rutgers Univ., New Brunswick, N.J. Lines: 21 In article <47215@apple.Apple.COM>, das@Apple.COM (David Shayer) writes: > Performance on RISC machines (and CISC machines, for that matter) often > has much more to do with the quality of the compiler than the speed of > the processor. Unless you knew that the compiler used on each machine > produced code of comparable quality, I don't you think have much of a > benchmark. While that usually may be true, when we're talking about the same application on different platforms, it's a different story (plus, Mathematica has its own interpreter for equations and programs, and compiler optimizations may not account for much performance differences except in I/O and user interface code. And, of course, in the efficiency of the interpreter. $-) ) Talking about theoretical differences in compiler optimization is fine, but in the case of Mathematica, we have a real-life example of an application which is considerably faster on the NeXTstation than the SS1+. I'd like to see some Mathematica benchmarks on the IIfx (if the bugs in the Mac version were ever fixed). -Declan McCullagh / Independent NeXT Developer