Path: utzoo!attcan!utgpu!jarvis.csri.toronto.edu!rutgers!ucsd!ucbvax!agate!stew.ssl.berkeley.edu!johnf
From: johnf@stew.ssl.berkeley.edu (John Flanagan)
Newsgroups: comp.sys.amiga.tech
Subject: Re: Shadow RAM
Message-ID: <1989Nov14.174952.8253@agate.berkeley.edu>
Date: 14 Nov 89 17:49:52 GMT
References: <42550@bu-cs.BU.EDU>
Sender: usenet@agate.berkeley.edu (USENET Administrator;;;;ZU44)
Distribution: comp
Organization: University of California, Berkeley
Lines: 23

In article <42550@bu-cs.BU.EDU> bear@bu-pub.bu.edu (Blair M. Burtan) writes:
>Does anyone think that Shadow RAM would be effective on an Ami?
...
>Basically this concept takes
>routines from ROM and places them in RAM.  The claim is that RAM
>is faster than ROM.  Would it be possible/worth the effort to
>take some of the graphics primitives from RAM and store them in ROM?

This has been done, and it is definitely worth it.  RAM is not inherently
faster than ROM, but if you copy the Kickstart ROMS into fast 32-bit RAM,
you get quite a performance increase over the 16-bit ROMS.  Dave Haynie's
SetCPU will do this if you have an A2620, and Brad Fowles' LUCAS/FRANCES
board combination will also do this.  I have the LUCAS/FRANCES combination,
and there is easily a 2- to 4-fold increase in speed in the system drawing
routines when the ROMs are remapped.

John


John Flanagan				Space Sciences Laboratory
johnf@sag4.ssl.berkeley.edu		University of California
(...!ucbvax!sag4.ssl!johnf)		Berkeley, CA 94720
Manners Maketh Man.			(415) 643-6308