Re: Model Instances
by Richard Van Stone · in Torque Game Engine · 12/11/2004 (4:49 pm) · 5 replies
I want to know how instancing works in torque. Basically....every time you load head model 1 it takes up memory. If I have say ten copies of model head 1 (as instances) does that consume the same amount of memory as having ten seperate head models (head 1 through head 9)? The second part of the question would be, if I gave each instance of model head 1 a seperate texture, is that the same amount of memory as say having head 1 through head 9 with those same textures?
About the author
#2
12/11/2004 (9:28 pm)
Ok, but how do I still track the memory useage for each model...as in the example I gave?
#3
12/12/2004 (11:44 am)
Um... Load your mission and see how much memory windows says you're using? For better accuracy, see how much memory Torque's memory manager says you're using...
#4
12/12/2004 (11:49 am)
I've never used the memory manager before, does it break down it's memory use per model?
#5
12/12/2004 (2:23 pm)
No, but if you take a recording of the amount of memory used before and after you load the shape you should get pretty useful numbers. These naturally don't correlate to usage of VRAM.
Associate Kyle Carter
"Instancing" in the sense of "only load assets once" is, naturally, supported. The ResourceManager implements this functionality. Textures are loaded and tracked seperately of geometry.