Path: utzoo!utgpu!news-server.csri.toronto.edu!cs.utexas.edu!uunet!mcsun!inria!bull.bull.fr!cediag!Patrick.Hayes From: Patrick.Hayes@cediag.bull.fr (Patrick Hayes) Newsgroups: comp.sys.mac.misc Subject: Re: Surely a IIFX blows an Amiga 3000 away, right? Message-ID: Date: 18 Oct 90 15:03:35 GMT References: <5046@crash.cts.com> Sender: news@bull.bull.fr Organization: Cediag Bull S.A., Louveciennes, France Lines: 40 In-reply-to: seanc@pro-party.cts.com's message of 15 Oct 90 20:16:04 GMT In article <5046@crash.cts.com> seanc@pro-party.cts.com (Sean Cunningham) writes: >Certain benchmarks you quoted would most deffinatly blow away a stock Amiga >3000, or A2500/30...you're comparing a 40MHz '030 to a 25MHz '030. Which is exactlty the point he was making. A stock fx DOES blow away a 3000, not the inverse as others have asserted. > A GVP or >Supra equipped A2x00 would leave the ][fx in the dust, as would a ][, ][cx, or >][x with a Daystar Digital accelerator. Which is outside the scope of the original posting, for one thing, and has little meaning especially between machines which do not run the same software. Comparing benchmarks means little, it's Application speed that counts. Benchmarks using typical applications and workloads have shown the fx to be a much better balanced machine than an accellerated IIx, and the improvements elsewhere than in the CPU clock speed give the fx a better rating in all the evaluations I've seen so far. >Where the 25MHz machines still blow any Mac away, regardless of clockspeed, is >in areas dealing with system performance across the buss, particularly when >dealing with memory. The ][fx has DMA for SCSI, but you'll have to wait to >take advantage of that. Ahem, I'm already using that DMA, using software that has been selling for months. It's called A/UX 2.0. Please don't make incorrect global generalizations. A/UX may be a bit over the head of the majority of mac users, but it DOES exist. In addition, I remember a few articles form someone at apple which gave numbers, and showed that for non-preemtive multi-tasking, DMA is actually SLOWER than using the CPU to move memory. Has anyone archived these articles? Pat -- +-------------------------------+-----------------------------------------+ | Patrick Hayes | EMail : Patrick.Hayes@cediag.bull.fr | | BULL CEDIAG | or hayes@bull.fr | | 68, Route de Versailles | or ...!mcvax!inria!bullfr!hayes | | F-78430 Louveciennes FRANCE | Tel : (33 1) 39 02 49 55 | +-------------------------------+-----------------------------------------+