home *** CD-ROM | disk | FTP | other *** search
- Path: sparky!uunet!noc.near.net!hri.com!spool.mu.edu!uwm.edu!cs.utexas.edu!torn!nott!bnrgate!bmerh85!bmerh85!hamish
- From: Hamish.Macdonald@bnr.ca (Hamish Macdonald)
- Newsgroups: comp.sys.amiga.hardware
- Subject: Re: 040 vs Rocket Launcher
- Message-ID: <1993Jan23.193715.27724@bmerh85.bnr.ca>
- Date: 23 Jan 93 19:37:15 GMT
- References: <73436@cup.portal.com> <weigner.3310@rcshq.adsp.sub.org>
- <1993Jan18.190358.15739@bmerh85.bnr.ca>
- <weigner.3504@rcshq.adsp.sub.org>
- Sender: news@bmerh85.bnr.ca (Usenet News)
- Organization: Bell Northern Research
- Lines: 24
- In-Reply-To: weigner@rcshq.adsp.sub.org's message of 22 Jan 93 11:12:58 GMT
-
- >>>>> On 22 Jan 93 11:12:58 GMT,
- >>>>> In message <weigner.3504@rcshq.adsp.sub.org>,
- >>>>> weigner@rcshq.adsp.sub.org (Jan Weigner) wrote:
-
- Jan> If you write optimised 68040 code (floating point, caches and
- Jan> using one or two new commands) it will definetely not run on a
- Jan> 68030. And 68040 optimised packages for ultimate performance are
- Jan> already coming righ now. There is a 68040 optimised version of
- Jan> Lightwave which is more than 10 times faster running on 28MHz
- Jan> 68040 card than on a Amiga 3000 (68030 25MHz). Also REAL3D is
- Jan> coming in 68040 optimised version being currently 8 times quicker
- Jan> than an Amiga 3000 running it on a 28MHz Fusion-Forty 68040
- Jan> board.
-
- I'm not certain about this, but I'm pretty sure that the 68040 does
- not change the user programming model of the 680X0 as compared to the
- 68030.
-
- I suspect that these new versions of these programs just have 2
- differences:
-
- 1) Code ordered to keep the pipeline filled.
- 2) Different selection of instructions which increase performance on
- an '040
-