Instancing or merging - performance cost

Development Platform & Builds XR Development
, , , , ,
1

Suppose 1 have 3 basic modular meshes.
With those meshes I’m building an entire space station.

When I’m done building should I:

  1. merge all meshes together to create one (merged) mesh for the spacestation ?
  2. leave the meshes as they are (group them in a Blueprint for convenience) and count on UE4’s instancing to deliver the most performance friendly solution

Which one of these two solutions would yield the best results in terms of polycount/performance (every poly counts (lol) – I’m devving for VR)

2

I’d say make large merged modular chunks from basic meshes instead of merging everything into a single mesh.

3

Thanks for chiming in. I guess the crux of the question remains unanswered. What would be the advantage between merging and not merging (not merging → just grouping everything in a blueprint). Wouldn’t merging (even in “larger chunks in stead of merging everything”) mean I’m creating a “new” mesh , in stead of working with nothing but instances and not creating anything new)

4

The advantage of merging would be a reduction of draw calls at cost of additional mesh memory.