[mythtv-users] New Nvidia drivers announced

[Jarod C. Wilson]

-----BEGIN PGP SIGNED MESSAGE-----
Hash: SHA1

On Wednesday 24 December 2003 06:57, Harry Orenstein wrote:
> From: "Jarod C. Wilson" <jcw at wilsonet.com>
>
> >>What type of CPU, memory and card do you have?
> >
> >Okay, so those numbers were on a dual Athlon MP 2000 system with 1.5GB of
> >RAM and a GeForce 4 Ti 4400. But I've seen numbers in 3000 range on an
> >Athlon XP 1700 w/a GF4MX and 512MB of RAM (DDR), if I recall correctly. I
> >could be mistaken, but James was pulling 1800 on a 2-year old laptop w/a
> >GF4MX...
> >
> >>I've got a Duron 800, 320MB PC100, NVidia FX 5200.
> >
> >AGP slot speed? Perhaps that is a bit of a bottleneck...
> >
> >Okay, now I read your next message. 2000FPS w/the 4363 drivers. That's
> > more like it. Sadly, nVidia seems to keep going downhill, post-4363...
> > I've got some work to do testing the 5328 drivers this week myself, but
> > the initial results make me think I'll stick to 4363.
> >
>
> Just wanted to jump in an ask a slightly more OT question:
>
> I have an Athlon XP 2000+ w/ GF4MX440se running 4363 and
> I get approx. 1800 FPS on my CRT (1024x800 32bpp) and
> 2400 FPS on my TV (800x600 24bpp).  My TV is set up with
> XvMC if that makes a difference.

Are those numbers from the same machine while connected to both of those 
displays at once? I know some facets of the nvidia driver only work on the 
primary display, but I'm not sure how that applies to OpenGL.

> Does resolution affect glxgears?

I'm not certain, since it just runs in a little window. I suppose it might 
have a bit of an impact.

> If so, what resolution are your
> numbers based on?

1600x1200

> What possible other factors might affect
> this benchmark?

Maximizing the gears versus running them in the default size window would have 
an impact, and that would definitely show lower numbers at higher 
resolutions.

> I am just trying to get a handle on how to use
> glxgears for a relative comparison and still compare "apples to
> apples".

Yeah, the maximized versus non-maximized one would cause major skew...

- -- 
Jarod C. Wilson, RHCE

Got a question? Read this first...
     http://catb.org/~esr/faqs/smart-questions.html
MythTV, Fedora Core & ATrpms documentation:
     http://wilsonet.com/mythtv/
MythTV Searchable Mailing List Archive
     http://www.gossamer-threads.com/archive/MythTV_C2/
-----BEGIN PGP SIGNATURE-----
Version: GnuPG v1.2.3 (GNU/Linux)

iD8DBQE/6fgtln8CX+oYdJURAp/lAKDqnfuLot6gKgLtpzU9s7Vx/MgoZgCcDMjD
JKAXGL1BMl0hq/shmfQI/8M=
=PQE/
-----END PGP SIGNATURE-----