In article <38698759.90BDCD6B@pacbell.net>, lin### [at] povrayorg 
wrote:

> Benchmark testing generaly relies on the amount of time it takes
> to render a given scene. If one were to design a scene that was
> parsing intensive rather than render intensive how well would it
> evaluate a systems performance ?
> 
> For example I designed this scene to take exactly 30 sec. to parse
> on my system. There is a camera added simply to ensure that it was
> not pointed at any object and there are no lights added so that I
> rendered a black screen only. Render time at 640x480 with aa = 0.3
> was only 8 sec. which is inconsequential. A pigment was added to
> the object to ensure that no CPU cycles were spent on sending no
> pigment warnings messages to the message window for each object
> created by the scene code.
> 
> camera {location < 0, 0, -3 >look_at  z*-6}
> 
> #declare A = 0;
>   #while (A<29500)
>     sphere{<0,0,A>,1 pigment{rgb 1}}
>   #declare A=A+1;
> #end

With the official 3.1g.r2, cache set to maximum:
Time For Parse:    0 hours  0 minutes  13.0 seconds (13 seconds)
Time For Trace:    0 hours  0 minutes   8.0 seconds (8 seconds)
    Total Time:    0 hours  0 minutes  21.0 seconds (21 seconds)

With MacMegaPOV:
Time For Parse:    0 hours  0 minutes  14.0 seconds (14 seconds)
Time For Trace:    0 hours  0 minutes   1.0 seconds (1 seconds)
    Total Time:    0 hours  0 minutes  15.0 seconds (15 seconds)

My computer is a Power Macintosh G3 266 "beige" desktop, with  96MB RAM, 
booting from an external SCSI 9.1GB LaCie hard drive. The OS I ran these 
tests under is Mac OS 8.6(I haven't gotten Mac OS 9 yet, and have been 
too busy to do anything with LinuxPPC).

I would suggest a scene file that does some more complex calculations 
instead of just creating a bunch of objects.

-- 
Chris Huff
e-mail: chr### [at] yahoocom
Web page: http://chrishuff.dhs.org/

Post a reply to this message