Path: utzoo!utgpu!jarvis.csri.toronto.edu!cs.utexas.edu!tut.cis.ohio-state.edu!rutgers!cbmvax!daveh
From: daveh@cbmvax.commodore.com (Dave Haynie)
Newsgroups: comp.sys.amiga.hardware
Subject: Re: (A) Amy 500 -> 2000? (B) 68040 vs. gfx coprocessor
Message-ID: <9588@cbmvax.commodore.com>
Date: 2 Feb 90 22:51:33 GMT
References: <633@xdos.UUCP> <9527@cbmvax.commodore.com> <640@xdos.UUCP>
Reply-To: daveh@cbmvax.cbm.commodore.com (Dave Haynie)
Organization: Commodore, West Chester, PA
Lines: 54

In article <640@xdos.UUCP> doug@xdos.UUCP (Doug Merritt) writes:
>In article <9527@cbmvax.commodore.com> mks@cbmvax.cbm.commodore.com (Michael Sinz - CATS) writes:

>Give me some credit for knowing that multitasking is important! But that's
>not the point. The point is that the 040 is supposed to average 1.3
>instructions per cycle. There are some questions that come to mind as
>a result: how much bus bandwidth is wasted by the processor that is
>even available to be used by a coprocessor??? 

Wasted?  One might ask "how much of the available bus bandwidth is used by
the processor", but you'd certainly hope it doesn't waste any.  In any
case, I recall that the 68030 uses about 50% of its available bus bandwidth,
leaving the remaining 50% available for other stuff on the same bus.  With
that mongo cache size, I expect the '040 to do much better.  Apparently, so
does Motorola; unlike the '030, the '040 only asks for the bus when it needs
the bus, more like the way hard disks work on the Amiga now than a normal
Motorola CPU.

But most of the coprocessors used in Amiga systems get an additional benefit
by sitting on alternate busses.  So they don't interfer with the CPU at all
unless the CPU needs access to their bus.

>And whether that particular system design would be cost effective in 
>Commodore's market.

That's one reason I think Agnus-like blitters will be with us for some time.
They cost significantly less than anything else you might throw into each
system to perform the same graphics work.

>There's nothing that says that vertical integration to the extent of
>maintaining your own chip foundry is always going to be the most cost
>effective of all possible choices!

Cost is a factor, but not the only factor.  Given the chip foundry, you
design computers differently at Commodore than you might elsewhere, where
the best thing you might have access to is an outside gate array fab.

>That's exactly the point. Considering such issues, it makes sense to
>*buy* 0.8 micron chips. Possibly even for coprocessors.

Of course it does.  But no one's proposing Commodore throw out the CPU
or the DRAM and start making our own.  There's lots of other stuff that
makes much more sense if we make it ourselves.  Making the parts internally
also saves a boatload of time and effort, not to mention NREs and other
development costs.

>Doug Merritt		{pyramid,apple}!xdos!doug
>Member, Crusaders for a Better Tomorrow		Professional Wildeyed Visionary


-- 
Dave Haynie Commodore-Amiga (Systems Engineering) "The Crew That Never Rests"
   {uunet|pyramid|rutgers}!cbmvax!daveh      PLINK: hazy     BIX: hazy
                    Too much of everything is just enough