From: dempson@actrix.gen.nz (David Empson)
Newsgroups: comp.sys.apple2
Subject: Re: 20MHz 65C816 (IIGS CPU) in production: used in C= accelerator
Date: Tue, 17 Dec 1996 21:00:55 +1300
Organization: Empsoft
Message-ID: <19961217210055195185@dempson.actrix.gen.nz>
References: <58psbf$10cu@uni.library.ucla.edu> <19961214193500.OAA01702@ladder01.news.aol.com> <58v1cq$rkd@darla.visi.com> <19961215075801.CAA23102@ladder01.news.aol.com> <C121596175712.8919@darkcastle.ecore.fnet.org>

Dark_Dude <Dark_Dude@darkcastle.ecore.fnet.org> wrote:

> In Article <19961215075801.CAA23102@ladder01.news.aol.com>,
> <rubywand@aol.com> wrote:
> 
> 
> >     It has already been observed that, as a 20MHz accelerator running
> >current GS wares, New GS would not greatly improve speed over accelerators
> >now available. Still, there will be some improvement. Getting to run
> >current wares on a 20MHz accelerator is a decent up-front benefit.
> 
> 
> We could gain some SERIOUS speed improvements! The buses are now running at 1
> Mhz. Make these, e.g. 10 or 20 Mhz, and see what it can do.

Correction: the I/O bus and slow memory interface (including video
buffers) now runs at 1 MHz.  This could not be changed.

What could be changed would be running all of the fast RAM (the main
128K to 8MB) at the full speed of the accelerator.  The accelerator
could also suck the ROM into on-card RAM and run it at full speed.

This means that code which makes heavy use of I/O or writes to video
memory would not gain much speed improvement, but everything else (which
is an awful lot) would be significantly faster.

-- 
David Empson
dempson@actrix.gen.nz
Snail Mail: P.O. Box 27-103, Wellington, New Zealand