From: elric AT wheel DOT dcn DOT davis DOT ca DOT us (Jeffrey Taylor)
Newsgroups: comp.os.msdos.djgpp
Subject: Re: FYI: performance of float vs. int
Date: 7 Feb 1997 17:43:15 GMT
Organization: Davis Community Network - Davis, California, USA
Lines: 14
Message-ID: <5dfpjj$lqt$1@mark.ucdavis.edu>
References: <199702061945 DOT LAA16451 AT netcom4 DOT netcom DOT com>
NNTP-Posting-Host: wheel.dcn.davis.ca.us
To: djgpp AT delorie DOT com
DJ-Gateway: from newsgroup comp.os.msdos.djgpp

Marc Singer (elf AT netcom DOT com) wrote:
<snip>
: I have been working on an implementation of a DCT (discrete cosine
: transform) and have found the suprising results that the fixed point
: integer version is slower than the floating point version on both on a
: 486 and on a P5.  I am not claiming that this will always be the case,
: but I was astonished to find that the fixed point version was twice as
: slow as the floating point implementation.  I intend to look into this
: more and will post again when I have a conclusion.
: 

I have the opposite experience, on a P90 a float inverse DCT increases the
run time by 10% over a fixed point inverse DCT.  The inverse DCT is about
half of the computation, so the real increase is more like 20%.