home *** CD-ROM | disk | FTP | other *** search
- Path: combo.ganesha.com!peterk
- From: peterk@combo.ganesha.com (Dr. Peter Kittel)
- Subject: Re: CHIP RAM speed test resul
- Distribution: world
- Newsgroups: comp.sys.amiga.programmer
- Reply-To: peterk@combo.ganesha.com
- References: <4j6jv0$1im@serpens.rhein.de> <5827.6659T112T770@mbox.vol.it> <1996Apr2.234528.8971@scala.scala.com> <4k1kk3$i2q@sunsystem5.informatik.tu-muenchen.de>
- Message-ID: <peterk.0mvy@combo.ganesha.com>
- Date: 13 Apr 96 22:22:47 MEZ
- Organization: Private Site
-
- In article <4k1kk3$i2q@sunsystem5.informatik.tu-muenchen.de> fischerj@Informatik.TU-Muenchen.DE (Juergen "Rally" Fischer) writes:
- >
- >BTW imho it should cost almost nothing to add a A3000-alike chipmembuffer,
-
- Oops, that Amber chip was one of the most expensive parts in the A3000!
- No, not at all like "cost almost nothing".
-
- >did you do it in walker ?
-
- No. Too expensive. We have some of the functionality in the AA chips,
- remember, so Amber was only needed for some very special cases, read
- games and c0der demos, for practically all other situation mode
- promotion (and in worst case some similar tool from AmiNet) does the job.
-
- >: >It would be sufficient to modify the interface CPU <->chipram<-> Alice.
-
- To explain it: This would mean a *complete* redesign of the AA chipset.
- And as Paula is still in some antique technology, it's not like just
- transferring the old design from 1.0 to 0.35 um, but to reinvent it
- from scratch...
-
- --
- Best Regards, Dr. Peter Kittel // Visit http://www.amiga.de
- Private Site in Frankfurt, Germany \X/ Email to: peterk@combo.ganesha.com
- Employed at AMIGA Technologies GmbH in Bensheim, Germany
-