Nanite, is it better to have lots of meshes or merge them down?

Nanite, is it better to have lots of meshes or merge them down? I am building a city environment… and im curious if vending machines can literally have every single can a seperate mesh inside that vending machine… or if its better to merge them. They arent very low poly, but not exacty high either. The difference memory wise would be 580 kb vs maybe 120 kb seperated as seperate objects. Every object uses the same material. So draw calls wont be an issue since its nanite im assuming.

I’m pretty sure the whole point of Nanite is to have an incredible amount of detail, and leave it to the system to manage it.

It will automatically change the resolution of objects, depending on viewing distance.

correct… but im not talking about the polycounts themselves now for performance. Memory is still an issue i would assume. So the question is … having 30 seperate cans in a vending machine… or merging them all down to 1 mesh… No real need to merge them down and saves memory if i can just keep them seperate… but i dont know what type of limitations im looking at for an entire city with bottles seperate in vending machiens lol…

What im saying is i dont know if it hurts performance down the road to have them all seperate.

If you’re really inquisitive, I’d say just try it both ways. But I think you’ll find keeping everything separate is better, because it gives the engine a chance to economize on meshes, textures and materials. If everything was merged, I don’t know if it could do that so easily.