[Question] Ray Intersection hardware resources

I’m using ray intersections for some collision checks, which works quite well.
I just wanted to ask: If there are more objects/triangles in a scene, would a ray intersection need more and more resources? There
are 9 Map Tiles that are relevant for ray intersection, is it better to do 9 ray intersections to avoid intersecting the whole scene, or can I just easily intersect my world root node?

(i’m quite new to graphics progamming :wink: )
Thanks for your answers :slight_smile:

It should use the bounding shape for early dismissal… so there is no reason to do 9 separate ones. That way is likely to be even less efficient.

If you have a lot of other stuff in your scene that you definitely don’t want to intersect (like thousands of objects [which is bad anyway]) then you could put the tiles in a common parent and intersect that.

Also if you are already using bullet (the native one) for your scene, you can let bullet do ray tests - they use the broadphase optimisation bullet offers- pretty performant usually.