benchmark

[titi]

In mod menu there is a benchmark now in scenario sections.

Tomreyn and me tried it with 3.5.2 and svn - 2554.
our results:

me (  AMD Athlon(tm) 64 X2 Dual Core Processor 4200+
OpenGL renderer string: GeForce 9600 GT/PCI/SSE2/3DNOW!
OpenGL version string: 3.2.0 NVIDIA 195.36.24 )

3.5.2:  93 fps
svn:   87 fps

tomreyn:

Code: [Select]

$ awk -F= '/^DISTRIB_DESCRIPTION/ {print $2}' /etc/lsb-release
"Ubuntu 11.04"

$ awk -F: '/^model name/ {print $2}' /proc/cpuinfo | head -n1
  Intel(R) Core(TM)2 Quad CPU    Q9400  @ 2.66GHz

$ free -m | awk '/^-\/\+ buffers\/cache:/  {print $NF}'
6516

$ lspci -knn | fgrep -A 3 VGA
01:00.0 VGA compatible controller [0300]: ATI Technologies Inc Radeon X1650 XT (Primary) (PCIE) [1002:7291]
   Subsystem: Micro-Star International Co., Ltd. Device [1462:0730]
   Kernel driver in use: radeon
   Kernel modules: radeon

$ glxinfo | grep -E '^OpenGL (renderer|vendor|version)'
OpenGL vendor string: X.Org R300 Project
OpenGL renderer string: Gallium 0.4 on ATI RV560
OpenGL version string: 2.1 Mesa 7.10.2

$ xrandr | head -n1
Screen 0: minimum 320 x 200, current 1680 x 1050, maximum 8192 x 8192

v3.4.0: 20 FPS
v3.5.2: 25 FPS
r2554: 29 FPS

[titi]

my son: geforce 8600GT   , AMD-X2  5600

3.5.2:  114 fps
svn:     104 fps

[tomreyn]

My low voltage dual core laptop with Intel GPU :

Code: [Select]

$ awk -F= '/^DISTRIB_DESCRIPTION/ {print $2}' /etc/lsb-release "Ubuntu 11.04"
user1@schleppie:~$ awk -F: '/^model name/ {print $2}' /proc/cpuinfo | head -n1
 Genuine Intel(R) CPU           U7300  @ 1.30GHz

$ free -m | awk '/^-\/\+ buffers\/cache:/  {print $NF}'
3269

$ lspci -knn | fgrep -A 3 VGA
00:02.0 VGA compatible controller [0300]: Intel Corporation Mobile 4 Series Chipset Integrated Graphics Controller [8086:2a42] (rev 07)
    Subsystem: ASUSTeK Computer Inc. Device [1043:1862]
    Kernel driver in use: i915
    Kernel modules: i915

$ glxinfo | grep -E '^OpenGL (renderer|vendor|version)'
OpenGL vendor string: Tungsten Graphics, Inc
OpenGL renderer string: Mesa DRI Mobile Intel® GM45 Express Chipset GEM 20100330 DEVELOPMENT
OpenGL version string: 2.1 Mesa 7.10.2

$ xrandr | head -n1
Screen 0: minimum 320 x 200, current 1366 x 768, maximum 8192 x 8192

v3.5.0: 15 FPS
v3.5.2: 17 FPS
r2554: 18 FPS

Update: Later tests with the same configuration showed r2554 as well as v3.5.2 to also have 15 FPS. So basically there is no noticeable change across these versions, and the differences are probably due to variable external influences (e.g. background applications such as apt-daemon getting busy during one benchmark but not during the other, or due to different cache states or varying I/O stress).

[titi]

my results here are very constant and always reproducible.
I think they are maybe related to the new way of terrain rendering?

[tomreyn]

 Reproducible in terms of always the exact same FPS value on the same game version with the same hardware, software and game configuration?

[titi]

yes, if I not try to sabotate it 

[softcoder]

me: geforce 9800GT   , Intel Quad Core

3.5.2:  134 fps
svn:     119 fps

[will]

So, in summary, those with big cards are slightly slower, but they are still measuring the fps in 100's.

And those with low-end cards are slightly to noticeably faster, and inch upwards towards playable fps.

Sounds like a good tradeoff?