From:	DHGW39B AT prodigy DOT com (Mr. Eric Domazlicky)
Newsgroups:	comp.os.msdos.djgpp
Subject:	Re: FP vs int timings
Date:	9 Jan 1997 17:55:16 GMT
Organization:	Prodigy Services Company 1-800-PRODIGY
Lines:	17
Distribution:	world
Message-ID:	<5b3be4$rj6@usenetw1.news.prodigy.com>
References:	<pine DOT gso DOT 3 DOT 95 DOT 970105232426 DOT 17458a-100000 AT hops DOT cs DOT jhu DOT edu> <pine DOT sun DOT 3 DOT 95 DOT 970107112402 DOT 15288a-100000 AT mdjunior>
NNTP-Posting-Host:	innugap4-int.news.prodigy.com
To:	djgpp AT delorie DOT com
DJ-Gateway:	from newsgroup comp.os.msdos.djgpp

> non-predictable non-repeatable overhead at that), you are not really
> comparing the divide times, so your conclusion that there is no
> significant difference between dividing ints and dividing doubles is 
>not
> valid.  When I subtracted the program overhead time, I found that on 
>my
> p5-133, int divides appear to take roughly one-and-a-half times as 
>long as
> double divides.
FPU instructions will likely be faster on a Pentium when you don't 
access the results in the next few instructions because the FPU works 
in parallel on this chip.  So the real question is do u want to make 
your game or whatever have any possibility of running reasonably on say 
a DX4-100 which doesn't have a parrallel FPU...

Eric Domazlicky
dhgw39b AT prodigy DOT com

- Raw text -