rdm@cfcl.UUCP (11/19/87)
FLAME ON I have been following the SPARC MIPS controversy with some interest. There is some room left for doubt left by Mr. Mashey's figures, but he admits this freely, and shows his assumptions and methods freely. Sun's John Hime, on the other hand, first talked about "not having to respond to a Sunnyvale start-up", then recently complained publicly of a "smear campaign" being leveled at Sun. Really. GIGGLE ON I was therefore more than a little amused to receive the following Sun press release, which will no doubt make its way into at least some of the industry journals: The following announcement was sent out November 10, 1987. For further information, please contact, John Loiacono, Sun Microsystems (415) 691-6424 Rich Dickerson, Daisy Systems (415) 960-6674. DAISY SYSTEMS BOOSTS PERFORMANCE OF BOARD ROUTING SOFTWARE WITH NEW SUN-4 RISC SERVER MOUNTAIN VIEW, Calif., November 10, 1987 - Sun Microsystems, Inc. and Daisy Systems Corporation announced today a new high performance network server ... the new system utilizes Sun Microsystems' RISC- based Sun 4/260 deskside server ... this new product performs routing tasks at six to seven times the speed currently achievable ^^^^^^^^^^^^^^^^^^ on Digital Equipment Corporation's MicroVAXII. ... "The 10 MIPS performance of the XL Server enables ... ^^^^^^^ Does that make the MicroVax II a 1.5 MIPS box? I always thought it was more like 90% of a 780... GIGGLE OFF Really, Sun, I'm sure you have a wonderful chip there, and I'm even willing to believe that it may achieve 10 MIPS when you get it all together. I may even buy one of the things, though I'll probably wait for the ECL version. In the meanwhile, however, could you try to restrain your Mr. Hime a bit, and/or produce some believable, verifiable numbers? FLAME OFF -- Richard Morin, proprietor {hoptoad,ptsfa}!cfcl!rdm Canta Forda Computer Laboratory +1 415 994 6860 Full spectrum consulting services for science and engineering.
davidsen@steinmetz.steinmetz.UUCP (William E. Davidsen Jr) (11/23/87)
I haven't tested the Sun4 myself, but a general comment about the performance of computers which was flamed here recently: there was a reference to performance 6 to 7 times that available on a MicroVAX-II. This is more useful than MIPS figures. Performance on one program may not reflect the general performance of two machines. If the code is in some way less optimal for a Sun4 than a MVII, then the wire routing figures and the MIPS could both be real. -- bill davidsen (wedu@ge-crd.arpa) {uunet | philabs | seismo}!steinmetz!crdos1!davidsen "Stupidity, like virtue, is its own reward" -me