[PyMOL] graphics performance

2003-01-08 Thread gordon
Hi 
I recently upgraded from a geforce 2 to a geforce 4 ti4200, and I don't
really notice a performance improvment (running linux). I judge this
based on how smoothly large sized proteins can be rotated with spheres
on.
Just wondering why?

Thanks in advance
Gordon

-- 
Gordon Wells
Department of Biochemistry
University of Pretoria



Re: [PyMOL] graphics performance

2003-01-08 Thread Dirk Kostrewa
Dear Gordon,

I assume that you changed the default nv driver in your XF86config(-4) file 
to the Nvidia driver nvidia. I've replaced on my private Pentium III (733 
MHz) PC a GeForce2 GTS against a GeForce4 Ti4200 and I also observe less 
improvement than expected. Here are my observations:
1. Under Windows98SE with 3DMark2001 I get a ~50 % higher score, which is 
probably okay.
2. Under Linux I get a ~ doubled frame rate for the simple OpenGL benchmark 
program gears in full-screen mode and somewhat less than this in the 
Vulpine OpenGL Mark, which is fine (Nvidia driver 31.23).
3. But, puzzlingly with pymol and on-screen rendering of a complex secondary 
structure sketch (without ray-tracing) I only get a moderate increase (~ 10 
%, or so) of the movie frame rate compared to a Geforce2MX440 running on a 
Pentium III (450 MHz) in my office! Here, I would really expect a huge 
difference in frame rate. I also don't understand this. A possible 
explanation would be that pymol, upon automatic detection of the graphics 
card, internally changes the complexity of the calculations resulting in a 
slower but improved image?

Best regards,

Dirk.

On Wednesday 08 January 2003 09:40, gordon wrote:
 Hi
 I recently upgraded from a geforce 2 to a geforce 4 ti4200, and I don't
 really notice a performance improvment (running linux). I judge this
 based on how smoothly large sized proteins can be rotated with spheres
 on.
 Just wondering why?

 Thanks in advance
 Gordon

-- 


Dirk Kostrewa
Paul Scherrer Institut
Life Sciences, OSRA/007
CH-5232 Villigen PSI, Switzerland
E-mail: dirk.kostr...@psi.ch
Phone: +41-56-310-4722
Fax: +41-56-310-4556
WWW: http://www.sb.psi.ch





Re: [PyMOL] graphics performance

2003-01-08 Thread Jules Jacobsen
Hi Dirk

This might actually be a hardware affect rather than a software effect.
The GeForce4
is undoubtably quicker that the GeForce 2 bu there comes a time when the
actual CPU is lagging behind in the amount of data it can pass to the GPU-
this also includes other limitations such as the graphics apature size
setting in the bios, the amount of RAM available to the machine and the
speed of the AGP BUS. Putting a GeForce4 into a PIII means the GPU will be
idling a lot probably upgrading the motherboard, RAM and CPU will yield
improvements, better graphics cards most likely won't.

Jules

On Wed, 8 Jan 2003, Dirk Kostrewa wrote:

 Dear Gordon,

 I assume that you changed the default nv driver in your XF86config(-4) file
 to the Nvidia driver nvidia. I've replaced on my private Pentium III (733
 MHz) PC a GeForce2 GTS against a GeForce4 Ti4200 and I also observe less
 improvement than expected. Here are my observations:
 1. Under Windows98SE with 3DMark2001 I get a ~50 % higher score, which is
 probably okay.
 2. Under Linux I get a ~ doubled frame rate for the simple OpenGL benchmark
 program gears in full-screen mode and somewhat less than this in the
 Vulpine OpenGL Mark, which is fine (Nvidia driver 31.23).
 3. But, puzzlingly with pymol and on-screen rendering of a complex secondary
 structure sketch (without ray-tracing) I only get a moderate increase (~ 10
 %, or so) of the movie frame rate compared to a Geforce2MX440 running on a
 Pentium III (450 MHz) in my office! Here, I would really expect a huge
 difference in frame rate. I also don't understand this. A possible
 explanation would be that pymol, upon automatic detection of the graphics
 card, internally changes the complexity of the calculations resulting in a
 slower but improved image?

 Best regards,

 Dirk.

 On Wednesday 08 January 2003 09:40, gordon wrote:
  Hi
  I recently upgraded from a geforce 2 to a geforce 4 ti4200, and I don't
  really notice a performance improvment (running linux). I judge this
  based on how smoothly large sized proteins can be rotated with spheres
  on.
  Just wondering why?