Xref: utzoo rec.games.programmer:3396 comp.os.msdos.programmer:4601 comp.graphics:17196
Path: utzoo!utgpu!news-server.csri.toronto.edu!rpi!zaphod.mps.ohio-state.edu!swrinde!elroy.jpl.nasa.gov!decwrl!deccrl!bloom-beacon!bloom-picayune.mit.edu!ahodgson
From: ahodgson@athena.mit.edu (Antony Hodgson)
Newsgroups: rec.games.programmer,comp.os.msdos.programmer,comp.graphics
Subject: Re: Need hard core help doing some 3d optimizations
Message-ID: <1991Apr12.140949.26414@athena.mit.edu>
Date: 12 Apr 91 14:09:49 GMT
References: <27979@uflorida.cis.ufl.EDU> <1991Apr11.202429.29648@kuhub.cc.ukans.edu>
Sender: news@athena.mit.edu (News system)
Organization: Massachusetts Institute of Technology
Lines: 23

In article <1991Apr11.202429.29648@kuhub.cc.ukans.edu> 2fmlcalls@kuhub.cc.ukans.edu writes:
>> I just ran Turbo Profiler on some code of mine, and it looks like roughly
>> 90% of my CPU time is being shoved into one function.  Any help on the
>> optimizations would be really appreciated.
>> ... float xa, ya, za;
>
>I don't fully follow C, but I know float.  Go with scaled integers or longs and
>use integer division.  

Check out the latest Dr. Dobb's journal on using memory-mapped math
coprocessors.  They claim that float division is not faster than 386
integer division on some PC hardware platforms.

>As well, use scaled integers for your sin/co look=up
>tables (i.e. sin(45 degrees) = 707 rather than 0.707).  When you are through
>multiplying, do a div 1000 (or C equivalent of integer division).

If you go this route, don't scale in base 10;  scale in base 2 and use
shift operations instead of division when possible.


Tony Hodgson
ahodgson@hstbme.mit.edu