home *** CD-ROM | disk | FTP | other *** search
- Path: sparky!uunet!zaphod.mps.ohio-state.edu!usc!elroy.jpl.nasa.gov!ames!agate!darkstar.UCSC.EDU!ucscb.UCSC.EDU!urbndv8
- From: urbndv8@ucscb.UCSC.EDU (Bleeding Rivets)
- Newsgroups: comp.os.os2.misc
- Subject: Re: ati graphics ultra
- Date: 25 Jul 1992 02:46:41 GMT
- Organization: University of California; Santa Cruz
- Lines: 56
- Distribution: comp
- Message-ID: <14qfahINNp0g@darkstar.UCSC.EDU>
- References: <1992Jul24.1480.22900@dosgate>
- NNTP-Posting-Host: ucscb.ucsc.edu
-
-
- In article <1992Jul24.1480.22900@dosgate> "roger ramsey" <roger.ramsey@canrem.com> writes:
- >Man Kit Tang wrote on 07-22-92 regarding ATI GRAPHICS ULTRA:
- >
- >MK>Well, for an AT bus based coprocessed grahics card. Most of the
- >MK>advantage gain does come from reducing bandwidth requirement of the AT
- >MK>bus. The Mach 8 does some graphics primitive pretty well, but for that
- >MK>matter, a 486 is not bad either.
- >
- >I think you'll find that 1024x768x256 on a 386/33 with a Graphics Ultra
- >will be faster than a 486/33 with a plain dumb frame buffer SVGA card.
-
-
-
- Changing CPU's to one that is twice as fast might give you ~ 10% speed
- improvement or so, if you already have a CPU fast enough to keep the bus
- busy; i.e. from a 386/33 to a 486/33.
- This proves to me it's not processor power that makes the real difference.
-
- >
- >MK>So, one more time, the problem of HIGH resolution graphics
- >MK>performance of most PC compatibles with AT based controller is the
- >MK>limited bandwidth of the AT bus.
- >
- >From the point of view that, using your above example, the 486 would have
- >to calculate every pixel and then send it across the bus rather than a
- >series of instructions to the card, then yes the bus becomes a limiting
- >factor.
- >
- >Roj
- >
-
- Your original statement was that the purpose of coprocessed video cards
- was to take the load off the CPU. I find this hard to believe in light of
- the fact that even a much faster CPU won't improve graphics speed much.
- If the CPU were being taxed by calculating pixels, and this is what kept
- video speed down, then a 2x faster CPU should draw a screen 2x as fast,
- and this is definitely not the case.
-
- (The above assumes both CPU's are using a dumb video card; with a coprocessed
- video card the bottleneck (in some cases) might once again be the CPU, since
- the bus traffic would be so greatly diminished.)
-
- > ~ WinQwk 2.0 a#473 ~ Reality is for those who can't handle Star Trek.
- >--
- >Canada Remote Systems - Toronto, Ontario/Detroit, MI
- >World's Largest PCBOARD System - 416-629-7000/629-7044
-
-
- --
- "The possessed's mad speech is the higher wisdom of the world, since it is
- human...Why have we not yet acquired this insight in relation to the world
- of the free will? Because outwardly we are the masters of madness, because
- the insane are violated by us, and we hinder them from living according to
- their ethical laws...Now we must endeavor to overcome the dead point in our
- realtionship to insanity." --Wieland Herzfelde, 1914
-