Re: [Dri-users] Measuring 2d/3d performance

[<ign...@no...>]

Craig Sylla <cs...@gm...> dijo: 
 
> What would be a reasonably quick and accurate way of measuring 
> graphics performance under Linux? 
>  
> I have two systems and need to get relative performance numbers for them. 
>  
> Thanks! 
> Craig 
>  
>  
> ------------------------------------------------------- 
> SF.Net email is Sponsored by the Better Software Conference & EXPO 
> September 19-22, 2005 * San Francisco, CA * Development Lifecycle 
Practices 
> Agile & Plan-Driven Development * Managing Projects & Teams * Testing & QA 
> Security * Process Improvement & Measurement * http://www.sqe.com/bsce5sf 
> -- 
> _______________________________________________ 
> Dri-users mailing list 
> Dri...@li... 
> https://lists.sourceforge.net/lists/listinfo/dri-users 
>  
 
I only know glxgears, which is part of Mesa. Most probaby you have it 
installed in your system. It tests 3D performance, but I don't know how  
good is the measuring. I also would  like to know other tests. 
 
You can also install a graphics game in both computers. If your hardware is 
slow enough, you can count the number of screen redrawings in ten seconds 
and compare.  The  game gl-117 can be set to different levels of graphics 
quality and it displays a number of FPS that can be used to compare. Bzflag 
is another game that can be used to do this. 
 
Another thing you can do is to play a DVD movie while having a CPU monitor: 
as lower is the CPU load, higher is the graphics performance. 
 
Speaking about graphics performance, there is somethig I would like to ask: 
when testing with glxgears, people always reports a single value of FPS, but 
I get two FPS meassures: one with the gears window on top, an another bigger 
value when I put the terminal window on top of the gears window. What is the 
value people usually report? what is the meaning of both ways to measure?