From:	Ludvig Larsson <ludvig AT club-internet DOT fr>
Newsgroups:	comp.os.msdos.djgpp
Subject:	Re: FASTEST DOUBLE BUFFERING?
Date:	Fri, 15 Jan 1999 03:58:58 +0100
Organization:	Faas-Goldhart
Lines:	21
Message-ID:	<369EAEF2.4621@club-internet.fr>
References:	<76jocl$lf0$1 AT ns DOT mtu DOT ru> <19990113010351 DOT 01578 DOT 00009727 AT ng36 DOT aol DOT com> <77j9i4$umh$1 AT news5 DOT svr DOT pol DOT co DOT uk>
NNTP-Posting-Host:	toulouse-camichel2-201.club-internet.fr
Mime-Version:	1.0
X-Trace:	front1.grolier.fr 916369009 8216 194.158.122.201 (15 Jan 1999 02:56:49 GMT)
NNTP-Posting-Date:	15 Jan 1999 02:56:49 GMT
X-Mailer:	Mozilla 3.01C-CLUB (Win95; I)
To:	djgpp AT delorie DOT com
DJ-Gateway:	from newsgroup comp.os.msdos.djgpp
Reply-To:	djgpp AT delorie DOT com

Well, in my even more humble opinion:)

Even newer computers videomemory is rather much slower than the
processor(even if their memory exceeds the computer memory in speed),
ex:

on my amd300 a loop wich moves bytes or an inline rep stosl
that transfers information from memory to videomemory is
equally fast.

And, on older gr-cards bank modes might be faster because
it's hardware implemented(and no software workarounds) and
the other way around, fornewer cards. Thats the most
important, not a houndred clockcycles to switch bank.

I'm always trying to move as little data as possible, ram->ram and
worse ram->videoram.

This is only what I belive:)

Ludvig Larsson

- Raw text -