[comp.windows.ms] Graphics cards, 768x1024x256

tneff@bfmny0.BFM.COM (Tom Neff) (02/17/91)

In article <3523@d75.UUCP> woan@peyote.cactus.org writes:
>In article <1991Feb17.161516.9589@ux1.cso.uiuc.edu> mcdonald@aries.scs.uiuc.edu (Doug McDonald) writes:
>>A friend got, a couple of days ago, a new 33 MHz 486 PC with an Orchid
>>Prodesigner Plus card with a meg of memory. He set it up to run Windows in the
>>768x1024x256 color mode. It is amazingly fast. Far faster than my 20 MHz
>>386 machine running a Video 7 card at 600x800x16 colors. Screen redraws
>>when returning from a full-screen DOS session are almost instantaneous.
>>
>>Have other people tried a similar setup? At least for what I tried, with
>>this setup I could only call Windows "blazingly fast".
>>
>>Is this card the way to go?
>
>It is not a bad card, but you must understand that these Super VGA
>cards are just dumb frame buffers, so the observed performance, is
>largely due to a faster CPU/bus driving the card and perhaps a better
>written Windows driver. For really good performance at such high
>resolutions, you should be looking at those with hardware assist (i.e.
>TI340?0 graphics coprocessors); they have come down in price and are
>certainly close to the Orchid's price level. Then again, any ET4000
>based card will probably perform similarly to the Orchid board and
>they start at around $220 for a board with 1 MB (Swan and PC Brand).

Cheaper than that - street price on the Boca SVGA+ is $175 or so.

I would definitely say if you have a fast 386 or 486, get one.  It's so
cheap you can use it somewhere else if you decide to get a fancy
coprocessor board later.

Definitely use the "Generic ET4000" drivers floating around BBS's and
CIS, rather than the Boca-supplied ones.  The Generics are newer, faster
and fix a bug.

Keep in mind that some WinApps have trouble with 256 colors.  Also that
16 colors is still noticeably faster than 256, even on the ET4000.

We're still looking for NON-interlaced 1024x768x16/256 Windows drivers
for the ET, though.  Existing drivers are interlaced.
 
-- 
          PISTOL LICENSES                X\X\X   Tom Neff
    Like to shoot?  Why commute?         \X\X\   tneff@bfmny0.BFM.COM
-- West Side Rifle & Pistol Range, NYC   X\X\X   uunet!bfmny0!tneff

mcdonald@aries.scs.uiuc.edu (Doug McDonald) (02/18/91)

A friend got, a couple of days ago, a new 33 MHz 486 PC with an Orchid
Prodesigner Plus card with a meg of memory. He set it up to run Windows in the
768x1024x256 color mode. It is amazingly fast. Far faster than my 20 MHz
386 machine running a Video 7 card at 600x800x16 colors. Screen redraws
when returning from a full-screen DOS session are almost instantaneous.

Have other people tried a similar setup? At least for what I tried, with
this setup I could only call Windows "blazingly fast".

Is this card the way to go?


Doug McDonald

woan@nowhere (Ronald S Woan) (02/18/91)

In article <1991Feb17.161516.9589@ux1.cso.uiuc.edu> mcdonald@aries.scs.uiuc.edu (Doug McDonald) writes:
>A friend got, a couple of days ago, a new 33 MHz 486 PC with an Orchid
>Prodesigner Plus card with a meg of memory. He set it up to run Windows in the
>768x1024x256 color mode. It is amazingly fast. Far faster than my 20 MHz
>386 machine running a Video 7 card at 600x800x16 colors. Screen redraws
>when returning from a full-screen DOS session are almost instantaneous.
>
>Have other people tried a similar setup? At least for what I tried, with
>this setup I could only call Windows "blazingly fast".
>
>Is this card the way to go?

It is not a bad card, but you must understand that these Super VGA
cards are just dumb frame buffers, so the observed performance, is
largely due to a faster CPU/bus driving the card and perhaps a better
written Windows driver. For really good performance at such high
resolutions, you should be looking at those with hardware assist (i.e.
TI340?0 graphics coprocessors); they have come down in price and are
certainly close to the Orchid's price level. Then again, any ET4000
based card will probably perform similarly to the Orchid board and
they start at around $220 for a board with 1 MB (Swan and PC Brand).



--
+-----All Views Expressed Are My Own And Are Not Necessarily Shared By------+
+------------------------------My Employer----------------------------------+
+ Ronald S. Woan       woan@peyote.cactus.org or woan%austin@iinus1.ibm.com +
+ other email addresses             Prodigy: XTCR74A Compuserve: 73530,2537 +

mcdonald@aries.scs.uiuc.edu (Doug McDonald) (02/18/91)

In article <91371965@bfmny0.BFM.COM> tneff@bfmny0.BFM.COM (Tom Neff) writes:
>Cheaper than that - street price on the Boca SVGA+ is $175 or so.
>
>I would definitely say if you have a fast 386 or 486, get one.  It's so
>cheap you can use it somewhere else if you decide to get a fancy
>coprocessor board later.
>
>Definitely use the "Generic ET4000" drivers floating around BBS's and
>CIS, rather than the Boca-supplied ones.  The Generics are newer, faster
>and fix a bug.
>
>Keep in mind that some WinApps have trouble with 256 colors.  Also that
>16 colors is still noticeably faster than 256, even on the ET4000.
>
>We're still looking for NON-interlaced 1024x768x16/256 Windows drivers
>for the ET, though.  Existing drivers are interlaced.
> 
>-- 

How are these cards related to the ORchid Prodesigner Plus? That one
is definately running in non-interlaced 768x1024/256 mode 
in Windows. No flicker at all. I was under the impression that 
interlaced or non-interlaced depended on a jumper on the card, not
on any software, though that is just an impression.

Doug McDonald

rdippold@maui.qualcomm.com (Ron Dippold) (02/19/91)

In article <1991Feb17.161516.9589@ux1.cso.uiuc.edu> mcdonald@aries.scs.uiuc.edu (Doug McDonald) writes:
>
>A friend got, a couple of days ago, a new 33 MHz 486 PC with an Orchid
>Prodesigner Plus card with a meg of memory. He set it up to run Windows in the
>768x1024x256 color mode. It is amazingly fast. Far faster than my 20 MHz
>386 machine running a Video 7 card at 600x800x16 colors. Screen redraws
>when returning from a full-screen DOS session are almost instantaneous.
>
>Have other people tried a similar setup? At least for what I tried, with
>this setup I could only call Windows "blazingly fast".

I have an AMI 33/386 with an Orchid Prodesigner II (1 meg, 1024x768x256) and it
really blazes.  I tried it in 256-color mode, and then in 16-color mode because
you hear so much about how much slower 256-color mode is.  I didn't notice a
whole lot of difference, but moved it back to 16-color mode because nothing
I use really needs the 256-color mode.

If you add the video speedup driver it's even faster.  In fact, my Sony 1304
takes longer to switch into 1024x768 mode than it does to redraw my Windows
screen (assuming a simple screen, not too many windows).  Just a <click> and
I'm there.

I hear the Diamond SpeedSTAR is even faster, but I got such a good deal on\
the PDII that I wasn't about to turn it down.

rdippold@maui.qualcomm.com (Ron Dippold) (02/19/91)

In article <3523@d75.UUCP> woan@peyote.cactus.org writes:
>In article <1991Feb17.161516.9589@ux1.cso.uiuc.edu> mcdonald@aries.scs.uiuc.edu (Doug McDonald) writes:
>>A friend got, a couple of days ago, a new 33 MHz 486 PC with an Orchid
>>Prodesigner Plus card with a meg of memory. He set it up to run Windows in the
>>...
>It is not a bad card, but you must understand that these Super VGA
>cards are just dumb frame buffers, so the observed performance, is
>largely due to a faster CPU/bus driving the card and perhaps a better
>written Windows driver. For really good performance at such high
>resolutions, you should be looking at those with hardware assist (i.e.
>TI340?0 graphics coprocessors); they have come down in price and are
 
Actually, you won't see a huge increase in performance using just any old
TI34010 graphics board, because of the way Windows handles graphics.  I
believe it was PC Magazine that noted that they were only about 20% faster
than a good SVGA board.

If you're doing real graphics work, however, the TI boards shred!

rdippold@maui.qualcomm.com (Ron Dippold) (02/19/91)

In article <91371965@bfmny0.BFM.COM> tneff@bfmny0.BFM.COM (Tom Neff) writes:
>
>We're still looking for NON-interlaced 1024x768x16/256 Windows drivers
>for the ET, though.  Existing drivers are interlaced.

I would suggest that you give the Orchid Prodesigner II drivers a try, as they
are not interlaced in those modes.  I have heard from others who have used the
OPDII drivers with the generic ET4000 boards with no problem, but I don't know
if they were using that high a resolution.

woan@nowhere (Ronald S Woan) (02/19/91)

rdippold@maui.qualcomm.com (Ron Dippold) writes:
Ron> Actually, you won't see a huge increase in performance using just
Ron> any old TI34010 graphics board, because of the way Windows
Ron> handles graphics.  I believe it was PC Magazine that noted that
Ron> they were only about 20% faster than a good SVGA board.

Ron> If you're doing real graphics work, however, the TI boards shred!

Side note to whomever mail me directly about this and did not get a
reply, my mail gateway was hosed last weekend and things might have
bounced.

Anyway, the PC Magazine issue to look at is the February 26, 1991
edition on pages 218-219. The NEC Multisync Graphics Engine at
1024x768 and 256 colors trounced the nearest competing Super VGA board
running 1024x768 and 16 colors by roughly 50%. Similar results at
800x600x256 vs 800x600x16... Granted that none of the Super VGA baords
tested was an ET4000 based one, the V-Ram VGA which it was tested
against is usually considered to be one of the better performers.
This is not the 500% type speedups that occur in vector or whatever
based applications (CAD), but it is significant considering that even
with double the number of bits being shoved around the coprocessed
boards are running significantly faster than the standard Super VGAs.

Anyway, bottomline 340?0 cards seem to go for $600+. Also an ad in the
same issue for ATI's 8514/Ultra claims to run Windows "24 times faster
than Super-VGA cards" based on PC Magazines WINBENCH 1.1... Hmm...
--
+-----All Views Expressed Are My Own And Are Not Necessarily Shared By------+
+------------------------------My Employer----------------------------------+
+ Ronald S. Woan       woan@peyote.cactus.org or woan%austin@iinus1.ibm.com +
+ other email addresses             Prodigy: XTCR74A Compuserve: 73530,2537 +