🚀 go-pugleaf

RetroBBS NetNews Server

Inspired by RockSolid Light RIP Retro Guy

Article View: alt.games.descent
Article #1816

Re: CTPCI and speeding up video

#1816
From: ma... Mr A Green
Date: Wed, 07 Jun 1995 00:00
52 lines
3191 bytes
In article <3r358n$b...@malakor.kku.ac.th>,
	Kitt Tientanopajai <ki...@gear.kku.ac.th.> writes:
>> Someone on alt.games.descent mentioned a pgm called CTPCI, which
>> allows you to control several PCI/Neptune/Mercury chipset parameters,
>> from a DOS command line.  It helped me boost my memory subsystem 
>> performance a bit, but mainly it helped improve my DOS/Windows
>> video performance a great deal.  By turning off CPU to PCI Bursting,
>> using CTPCI's -B option, I was able to get the ATI Gfx Pro Turbo card to
>> run MUCH faster in any video mode that uses the "old style" SVGA/VGA modes, 
>> which run out of the < 1M address space (all DOS games currently out there
>> use the older "banked" modes, and the VGA 320x200 etc modes all
>> run out of this space).
>> 
>> The strange thing is, that apps which use the card in native mode,
>> using the linear aperature modes in the PCI address space (real high up
>> there, like at the 4016MB range I think), turning off PCI BURST makes
>> things much slower.  The OS/2 video driver for the ATI Mach 64 uses
>> the card in it "native" mode, so the -B option slows things up
>> for OS/2.  When using the VESA 2.0 test pgm VBETest, the linear
>> aperature VESA 2.0 modes slow down by more than 50% when you
>> turn off CPU to PCI Bursting. So theoretically games that 
>> come out using the advanced VESA 2.0 linear gfx modes, will
>> be a lot faster if you use the +B (turn on CPU to PCI BURST)
>> with the ATI Mach 64 card. For now under DOS/WIN -B is best.
>> 
>> Windows 3.11 and the ATI Mach 64 drivers actually SPEED UP by quite
>> a bit when you turn off CPU to PCI BURST..so it looks like the Windows
>> 3.11 stuff DOES NOT use the ATI card in its fastest (linear aperature)
>> mode...seems like it should.
>> 
>> Oh well, for now I am happy, my DOS video games are  A LOT faster,
>> Windows is a bit faster, 3DBench went from 71 fps to over 88 fps..
>> and I can still keep PCI BURST on for OS/2 so it does not lose anything.
>> 
>> Cards like the Hercules w32p's, the Hercules ARK2000's and the s3 864/964
>> based cards seem to work fine in all modes when the BURST is on..it is
>> only the ATI card that seems to love BURST for the linear aperature
>> modes, and hate it for "normal/oldstyle" modes.
>> 
>> -Roger
>> --------------------------------------------------------------
>> NET: uz...@crash.cts.com
>> *WINDOWS from the folks who brought you EDLIN
>
>Umm...How can I find it ?...Where can I FTP it ?

I want it too. Can somebody post it here??? or at least give the ftp
site its lying at.
Oh, also, EDLIN wasn't wriiten by microsoft, but someother company. so there.
:)

Adam
Technical Details
Message-ID: <3r44u7$q4@lily.csv.warwick.ac.uk>
Path: mbox-mo1.pugleaf.net!unknown!not-for-mail
References: <D9KMp0.G2J@crash.cts.com> <3r358n$bk@malakor.kku.ac.th>

Show Full Headers
Complete Article Headers
From: ma...@csv.warwick.ac.uk (Mr A Green)
Subject: Re: CTPCI and speeding up video
Date: 1995/06/07
Message-ID: <3r44u7$q4@lily.csv.warwick.ac.uk>
X-Deja-AN: 103902280
references: <D9KMp0.G2J@crash.cts.com> <3r358n$bk@malakor.kku.ac.th>
organization: University of Warwick, Coventry, UK
mime-version: 1.0
newsgroups: comp.os.os2.advocacy,alt.games.descent,comp.sys.ibm.pc.games.action,comp.benchmarks
Content-Type: text/plain; charset=UTF-8
Content-Transfer-Encoding: 7bit