home *** CD-ROM | disk | FTP | other *** search
- Path: informatik.tu-muenchen.de!fischerj
- From: fischerj@Informatik.TU-Muenchen.DE (Juergen "Rally" Fischer)
- Newsgroups: comp.sys.amiga.hardware
- Subject: Re: Nemac IV on CV64
- Date: 3 Apr 1996 01:27:00 GMT
- Organization: Technische Universitaet Muenchen, Germany
- Distribution: world
- Message-ID: <4jsk54$i7q@sunsystem5.informatik.tu-muenchen.de>
- References: <haraldk.64.31444B59@sofus.hiof.no> <xS1t3MD42Dacz9@point70.people-s.people.de> <816.6657T148T1558@burst.demon.co.uk>
- NNTP-Posting-Host: hphalle5.informatik.tu-muenchen.de
- X-Newsreader: TIN [version 1.2 PL2]
-
- Lee Huggett (Lee@burst.demon.co.uk) wrote:
-
- : I read in the instructions that the fact that it is slow is due to the
- : programmers using a certain FPU instruction that isn't present in all the FPU
- : unit of the 68K series...
-
- : Its slow on my '060 + CV64 too...
-
- obviously this is slow software. Stop thinking "even my 060 is to
- little", please.
-
- I tested an mapping inner loop algorithm of a 386-25sx. it seems
- to have no icache, it was about 10% faster than my 020-14 with fastmem
- doing a similar loop.
-
- but: doom on 386-25 runs factors faster than textdemo57 (the
- fastest clone I know) on my 020-14.
-
- see ?
-
- You should cry for proper software. Just tell me a clone uses fpu
- and I tell you only someone who doesn't optimize for fast gfx
- would use floating point for current games.
-
- : Lee Huggett lee@burst.demon.co.uk
- : c9251123@zeus.hud.ac.uk
- ------------------------------------------------------------------------
- fischerj@Informatik.TU-Muenchen.DE (Juergen "Rally" Fischer) =:)
-
-