Path: utzoo!utgpu!watmath!clyde!att!rutgers!mailrus!tut.cis.ohio-state.edu!bloom-beacon!husc6!purdue!decwrl!vixie From: vixie@decwrl.dec.com (Paul A Vixie) Newsgroups: comp.sys.nsc.32k Subject: Re: Goodspeed GS-32 users, anybody out there? Message-ID:Date: 29 Nov 88 09:21:05 GMT References: <1036@majestix.liu.se> <1038@mina.liu.se> <52@gnome6.pa.dec.com> <67@sopwith.UUCP> Sender: vixie@decwrl.dec.com Organization: DEC Western Research Lab Lines: 21 In-reply-to: snoopy@sopwith.UUCP's message of 27 Nov 88 21:55:29 GMT [Snoopy] # Something doesn't compute here. For most things, a 32016/10Mhz box is # roughly the same as a VAX 750, not 1/2 of it. Have you run other benchmarks? # Dhrystones are a really poor benchmark, IMHO. If you ask it to calculate floating point stuff, it's about equal to a 750 without an FP750 (i.e., 750 running floating point emulation). If you ask it to sort integers, it's about 70% of a 750. If you ask it to do lots of procedure calls, it's about 70% of a 750. If you run a tight loop doing arithmetic on some registers, it's about 70% or 80%. If you put two users on both machines, running VI or JOVE, it's about half a 750. If you put four users on them, it's about 10% of a 750 on a per-user basis. Half a 750 seems about right to me after using both for extended periods of time. -- Paul Vixie Work: vixie@decwrl.dec.com decwrl!vixie +1 415 853 6600 Play: paul@vixie.sf.ca.us vixie!paul +1 415 864 7013