Benchmark: Riva TNT2 under Linux
David Gobbi
dgobbi at irus.rri.on.ca
Sun Aug 1 19:00:33 EDT 1999
I've applied the sphere-bench to some of the machines
in our lab, the results are listed below.
At the same time, I'd like to put in a plug for the nVidia TNT2
board -- for most of what we do, it is the fastest board we have.
It doesn't show in the sphere bench results (which test geometry
acceleration) but I've found that the texture-map acceleration on
the TNT2 is better than twice as fast as that on the SGI-320.
The new Matrox G400 is also hardware-accelerated under Linux,
if anyone wants to give it a try.
We do harware volume-rendering via texture maps, so texture mapping
is pretty much everything to us -- on the TNT2 a 128x128x128 volume
can render at 9fps in a 800x800 window, or about 30fps in a 400x400
window. This is with 16-bit textures, the drivers don't support
32 bit textures yet.
Riva TNT2 32MB (ELSA Erazor III) under 1x450MHz i686 (PII) Linux-2.0
Notes: 1) 16 bits (32 bits is unaccerated in this version of driver)
2) uses nVidia glx from their web site
3) performance should increase significantly with
XFree86-4.0 release and Direct-Rendering
4) full size of screen was 1800x1440 (Wow!)
Best results, summary :
- i686 running Linux 2.0.36
- Tcl/Tk 8.0.3
- OpenGL (visual is truecolor/truecolor/16)
- window was 400 x 400
1) 512x512 : 238.9 kpolys/s : [stripper]
2) 512x512 : 114.0 kpolys/s : []
3) 512x512 : 114.0 kpolys/s : [transparency]
4) 512x512 : 113.6 kpolys/s : [small_sphere]
5) 512x512 : 107.6 kpolys/s : [texture, transparency]
6) 512x512 : 107.5 kpolys/s : [texture]
7) 512x512 : 52.6 kpolys/s : [wireframe]
Cobalt, 1x450MHz i686 (PII) SGI 320 Visual Worstation, NT4.0
Notes: 1) recent OpenGL drivers (earlier drivers were slower)
2) 32 bits, screen size is 1600x1200
Best results, summary :
- intel running Windows NT 4.0
- Tcl/Tk 8.0.5
- Win32OpenGL (visual is truecolor/truecolor/32)
- window was 400 x 400
1) 512x512 : 4166.8 kpolys/s : [stripper]
2) 512x512 : 1397.6 kpolys/s : [transparency]
3) 512x512 : 1393.9 kpolys/s : [small_sphere]
4) 512x512 : 1373.1 kpolys/s : []
5) 512x512 : 1030.7 kpolys/s : [texture]
6) 512x512 : 1027.4 kpolys/s : [texture, transparency]
7) 512x512 : 686.3 kpolys/s : [wireframe]
ATI Rage 128 Fury, 1x450MHz i686 (PII), NT4.0
Notes: 1) 1600x1200 display, 32 bits
Best results, summary :
- intel running Windows NT 4.0
- Tcl/Tk 8.0.5
- Win32OpenGL (visual is truecolor/truecolor/32)
- window was 400 x 400
1) 256x256 : 432.9 kpolys/s : [stripper]
2) 256x256 : 188.8 kpolys/s : []
3) 256x256 : 188.8 kpolys/s : [transparency]
4) 256x256 : 187.3 kpolys/s : [small_sphere]
5) 256x256 : 187.1 kpolys/s : [texture, transparency]
6) 256x256 : 186.7 kpolys/s : [texture]
7) 256x256 : 163.3 kpolys/s : [wireframe]
- David
--David Gobbi, MSc dgobbi at irus.rri.on.ca
Advanced Imaging Research Group
Robarts Research Institute, University of Western Ontario
-----------------------------------------------------------------------------
This is the private VTK discussion list. Please keep messages on-topic.
Check the FAQ at: <http://www.automatrix.com/cgi-bin/vtkfaq>
To UNSUBSCRIBE, send message body containing "unsubscribe vtkusers" to
<majordomo at gsao.med.ge.com>. For help, send message body containing
"info vtkusers" to the same address. Live long and prosper.
-----------------------------------------------------------------------------
More information about the vtkusers
mailing list