From: John Stone (johns_at_ks.uiuc.edu)
Date: Wed Apr 11 2001 - 01:31:59 CDT

Dominik,
  As far as high-end cards go, I can't comment too much as we have
basically stuck with using the GeForce 2 line of cards in our lab.
I've read some reviews that indicate that the FireGL cards would
handily outperform the GeForce cards, however I have no direct
experience using them, and their Linux support was still "beta"
last I checked. They did look like an interesting possibility.
I doubt that the Quadro boards would give you much more VMD
performance than a GeForce would, but that's just based on the
benchmarks I've read out on the net, not on firsthand testing.
The FireGL boards are also somewhat more interesting for a high-end
solution as I believe they have full support for stereoscopic
display, whereas the GeForce boards do not. I don't think the Quadro
boards support stereo from what I remember, but this may depend on
which vendor you get the board from. As far as GeForce3, it looks
to me like the GeForce 3 will perform better for VMD, but the difference
may only be noticable when running in high-resolution display modes
such as 1280x1024 or 1600x1200, or with antialiasing enabled.
The GeForce3 _could_ become a very good card for VMD depending on
whether or not we are eventually able to find ways to use its unique
features to benefit VMD. At this time it is not clear whether we'll
be able to benefit significantly, but in any case it would be months
from now before we would have the time to create GeForce3-specific
rendering code for VMD.

In answer to your last question, the performance of the high-end
boards typically isn't wildly faster than the consumer-priced cards,
but they definitely offer better features for things like hardware
antialiasing, stereoscopic rendering, increased texture memory,
and acceleration of a larger subset of OpenGL...

Let us know if you have more questions on this stuff.

Thanks,
  John Stone
  vmd_at_ks.uiuc.edu

On Mon, Apr 09, 2001 at 04:09:45PM -0600, Dominik Horinek wrote:
> Dear John,
>
> thanks for the fast reply.
> Can you give me some recommendations about medium- and high priced
> accelerators, too? I've read something about Nvidia Quadra2, but these
> ones are very expensive. What about Geforce3?
> Is the performance so mucher better compared to the low-priced cards?
>
> Greetings,
>
> Dominik
>
>
> --
> please reply to one of the following addresses:
> horinek_at_eefus.colorado.edu
> horinek_at_gmx.net
>
>
> On Mon, 9 Apr 2001, John Stone wrote:
>
> >
> > Dear Dominik,
> > At the present time I'd recommend the NVidia GeForce family of
> > products as a low-cost option for Linux. You can pick up several
> > of the GeForce models for between $100 and $150, and they all work
> > pretty well on Linux, using the XFree86 4.x server that comes with most
> > of the current Linux distributions. The drivers are pretty easy to
> > install, and cards perform very well given what they cost. There are
> > higher-end "professional" OpenGL accelerators for Linux, as well as
> > commercial X servers. In our lab, we're using a large number of the
> > "GeForce 2 GTS" model cards, ours are sold by "WinFast", but they are
> > almost the same as the NVidia reference boards.
> >
> > Let us know if you have more questions about this.
> >
> > Thanks for using VMD!
> > John Stone
> > vmd_at_ks.uiuc.edu
> >
> > Let me know if you have more questions about this.
> >
> >
> >
> > On Mon, Apr 09, 2001 at 03:38:40PM -0600, Dominik Horinek wrote:
> > > Hi,
> > >
> > > I'd like to have recommendations for a 3d graphic card. VMD/ Linux should
> > > support it using OpenGL 3d acceleration. What prices do I have to expect
> > > for such cards.
> > >
> > > Thanks for any recommendations,
> > >
> > > Dominik
> > >
> > > --
> > > Dominik Horinek
> > > University of Colorado at Boulder
> > > reply to: horinek_at_eefus.colorado.edu
> >
> > --
> > NIH Resource for Macromolecular Modeling and Bioinformatics
> > Beckman Institute for Advanced Science and Technology
> > University of Illinois, 405 N. Mathews Ave, Urbana, IL 61801
> > Email: johns_at_ks.uiuc.edu Phone: 217-244-3349
> > WWW: http://www.ks.uiuc.edu/~johns/ Fax: 217-244-6078
> >

-- 
NIH Resource for Macromolecular Modeling and Bioinformatics
Beckman Institute for Advanced Science and Technology
University of Illinois, 405 N. Mathews Ave, Urbana, IL 61801
Email: johns_at_ks.uiuc.edu                 Phone: 217-244-3349              
  WWW: http://www.ks.uiuc.edu/~johns/      Fax: 217-244-6078