Relay-Version: version B 2.10 5/3/83; site utzoo.UUCP
Posting-Version: version B 2.10.2 9/18/84; site aum.UUCP
Path: utzoo!watmath!clyde!burl!ulysses!gamma!epsilon!zeta!sabre!petrus!bellcore!decvax!ucbvax!ucdavis!lll-crg!well!ptsfa!aum!freed
From: freed@aum.UUCP (Erik Freed)
Newsgroups: net.micro.amiga,net.micro.atari
Subject: Re: Reply to: Questions about ST vs. AMIGA
Message-ID: <414@aum.UUCP>
Date: Sun, 22-Dec-85 12:29:30 EST
Article-I.D.: aum.414
Posted: Sun Dec 22 12:29:30 1985
Date-Received: Wed, 25-Dec-85 03:21:11 EST
References: <342@unmg.UUCP> <840@h-sc1.UUCP> <352@well.UUCP> <382@pedsgd.UUCP> <425@amiga.amiga.UUCP>
Distribution: net
Organization: The Aurora Systems Bunch
Lines: 35
Xref: watmath net.micro.amiga:1264 net.micro.atari:2093

> In article <382@pedsgd.UUCP> bobh@pedsgd.UUCP (Bob Halloran) writes:
> >Granted, the graphics chips relieve the 68000 from dealing with the
> >display, as it must in the Mac, but the 68000 and the various co-processors
> >must still contend for access to memory.  YES, I/O can proceed independently,
> >etc., but only ONE of the processors can use memory at a time.  This
> >HAS to impact the throughput of the 68000.
> 
> This is true.  But don't forget that the Amiga interleaves its access
> to the bus, using every available cycle, even the odd ones, whenever
> possible.  All of the DMA, including the display processors, use the
> odd cycles.  This enormously cuts down the contention for the bus.
> The result of this is that the 68000 and other co-processors get to run 
> at full speed regardless of display processing.  Only when the
> display gets elaborate -- 640 across (either interlaced or not) and
> 4 bit-planes (16 colors onscreen) -- is there a significant lag
> in CPU performance.
>     The same goes for the other DMA devices:  sprite, disk, and audio.
> With a simple display, a few sprites and some simple tones playing,
> you can fill a disk buffer and still have truly minimal interference
> with the performance of the 68000.

I would still like to know why the Amiga even with a 68020 and fast memory
still is *significantly* slower than the ST. Is their a hardware type out their
who can look at the memory cycles and see why the current claims of "no cpu
slowing due to graphics chips" seems to not be proved by the real world. If
I had one I would do it myself. If possible could a Amiga hardware engineer
explain this disparity? It seems very strange. P.S. I know that the clock is
slightly slower.

-- 
-------------------------------------------------------------------------------
                           Erik James Freed
			   Aurora Systems
			   San Francisco, CA
			   {dual,ptsfa}!aum!freed