Relay-Version: version B 2.10 5/3/83; site utzoo.UUCP
Path: utzoo!mnetor!uunet!husc6!cmcl2!rutgers!ucla-cs!zen!hoser.berkeley.edu!bryce
From: bryce@hoser.berkeley.edu (Bryce Nesbitt)
Newsgroups: comp.sys.amiga
Subject: Re: New Custom Chips?
Message-ID: <3823@zen.berkeley.edu>
Date: Fri, 18-Sep-87 23:11:45 EDT
Article-I.D.: zen.3823
Posted: Fri Sep 18 23:11:45 1987
Date-Received: Sun, 20-Sep-87 12:14:24 EDT
References: <2948@hoptoad.uucp> <2345@cbmvax.UUCP> <2472@xanth.UUCP>
Sender: news@zen.berkeley.edu
Organization: Tubular Transport, Inc.
Lines: 46
Summary: Amiga 1000 possibilities... just build (another) tower.

In article <2472@xanth.UUCP> kent@xanth.UUCP (Kent Paul Dolan) writes:
>In article <2345@cbmvax.UUCP> grr@cbmvax.UUCP (George Robbins) writes:
>>In article <2948@hoptoad.uucp> slc@hoptoad.UUCP (Steve Costa) writes:
>>> [2 Meg of chip RAM?]
>>
>>Commodore has improved Amiga custom chips under development.	Packaging and
>>production volume considerations make it unlikely that these chips will be
>>available in A1000 compatible form.  [George R. Robbins]
>
>Sob.  Really hope you can change your minds on this one, George...

I'll bet someone could take the new chips and build a tower to stick them
into an Amiga 1000.  Assuming a hypothetical "Fat Agnus" pin compatible chip
with 1 or 2 Megabytes of CHIP RAM potential the logic to do this is
convoluted, but looks doable.  You would need to plug into the current Agnus
socket and run a clip or three down to the motherboard.  Most of the logic on
the board would be to unwind and/or selectively ignore the extra logic inside
Fat Agnus.  With FAST enough logic, timing margins should still be ok.	(BTW:
There was an obscure joke in that last sentence).  The rest of the board
would be an extra 512K or 1.5Meg of CHIP memory.

[BTW: I know a company that would like to sell this.  In fact, I've been
asked to the design it.  Now if only Commodore comes out with the part! :-) ]

The only real problem would be getting the parts OEM from Commodore. Should
not be too bad... Commodore has been getting back into OEM recently. (After
shutting MOS technology off from most of the world for a few years).

If the new chips will be in a Fat Agnus style package is matter of
speculation from this point of view.  It seems unlikely that one would go
through the trouble of expanding all 25 DMA channels, and then do it by only
one bit.  By the time everything is shuffled to make space 2 or 4 or more
bits are likely (even if all of those are not brought out of the package).
Perhaps the test line will do double duty as a memory select?	Nah.

Now if the chips George is typing about support 32 or 64 bit wide access
(that's CHIP width, not processor width), non NTSC video option, or straddle
CHIP/FAST memory for a 32 address bit system wide DMA facility, *then* it's
time to donate your Amiga 1000 to the local school or IBM-PC'ers anonymous
chapter.

 
|\ /|  . Ack! (NAK, ENQ, SYN)
{o O} . 
 (") 	bryce@hoser.berkeley.EDU -or- ucbvax!hoser!bryce
  U	How can you go back if you have not yet gone forth?