Path: utzoo!utgpu!news-server.csri.toronto.edu!cs.utexas.edu!samsung!zaphod.mps.ohio-state.edu!uwm.edu!ogicse!pdxgate!eecs!hal
From: hal@eecs.cs.pdx.edu (Aaron Harsh)
Newsgroups: comp.sys.amiga.advocacy
Subject: Re: A3000UX - Born to run UNIX SVR4
Message-ID: <1521@pdxgate.UUCP>
Date: 9 Feb 91 06:04:12 GMT
References: <1991Feb7.151106.4795@cc.helsinki.fi> <32530@auc.UUCP> <1512@pdxgate.UUCP> <11964@helios.TAMU.EDU>
Sender: news@pdxgate.UUCP
Reply-To: hal@eecs.UUCP (Aaron Harsh)
Organization: Roman Catholic Church
Lines: 59

In article <11964@helios.TAMU.EDU> n177ac@tamuts.tamu.edu (Daryl Biberdorf) writes:
>Since when does the processor alone determine the total capability
>of the machine?  We had an 030 NeXT in the integration lab at my
>last co-op assignment.  As long as you wanted to run ONE job at a time
>the machine's performance was excellent.  But try feeding a score file
>to the DSP (yes, teh sound *is* wonderful) while doing one or
>two other tasks and watch the whole thing grind to a halt (INCLUDING
>the music).

  Not on the new NeXT's.  On my lowest-end system, I can run Scoreplayer
while I'm switching between a word processor and a graphics-intensive game
and everything works fine.

>The NeXT's architecture must be such that the *CPU* is
>responsible for feeding data to the DSP (and, by analogy, to the
>other specialized chips).  With that kind of architecture, the NeXT
>will absolutely have to have an 040 to get any kind of performance
>at *all*.  I do not know all the details of the NeXT's innards, but
>after watching it react to my actions, I'd say that it has very little
>custom chips to handle DMA and I/O.  The 3000UX *does*.

  Here's a quote from some NeXT propoganda:
  "[the NeXT uses] two chips.  One contains 12 Input/Output processors.
each with direct access to meory; the other contains the circuitry needed
to manage the mass storage."

>I'd be willing to bet that the 3000UX is more than a match for the
>040 NeXT after having seen a 3000 running AmigaDOS (Gee is that fast)
>versus the NeXT choking under its version of Mach.  Putting the
>overhead of UNIX on the 3000 won't slow it down as much as
>that 030 NeXT did. 

  I haven't seen a 3000UX (but you haven't seen a NeXTstation, so I guess
we're even), but I'm sure it's slowed down an awful lot by UNIX.  Mach is
one of the fastest UNICES around, one of the reasons NeXT chose it, I'd
imagine.
  The thing that really slowed down the '030 NeXT was Display Postscript.
This has been pretty much fixed by the new machines.  Windows move fast,
text scrolls fast, GIF's load fast.  Admittedly, the NeXT isn't as fast
as an Amiga for graphics, but unless you're using virtual reality, or
a flight simulator, screen I/O won't be the bottleneck.

>When will everyone here who keeps griping about the price point that
>companies have to make a profit to SURVIVE!  Has it occurred to
>*anyone* that NeXT's low prices are a near-money-losing venture
>in order to get MARKET SHARE?!

  So what you're saying is that NeXT is offering low priced products
hoping that people will buy them?

>UNIX is not cheap to license.  I expect that Commodore is paying
>a LOT for the rights to SVR4.

  I don't think it's that much.  I'd guess <$100/machine.  AT&T shouldn't
be getting too greedy now that the DARPA-funded Mach project has removed
all the AT&T code from their OS.  (Free UNIX clones!!)

Aaron Harsh
hal@eecs.cs.pdx.edu