I was wondering, does UE5 render the inside of a closed environment, where light/photons won’t enter and how does UE5 handles it if the camera/viewport goes inside the closed environment, does UE5 continue to compute or stop the computation to increase GPU performance?
And what’s UE5 doing different than VALVE in render optimization?
ref link for cs1.6, how VALVE saves rendering in closed environment:
this is a complex topic, off course it does culling and the stuff you are asking. there are also volumes and options to give more priority to certain areas. take nanite for example, is the most advanced tech that we have in game design at the moment. The fact that you can use high poly models and you can avoid creating low poly versions manually, also you can avoid baking lights which needs carefully planned UVs. Man those were really annoying for a single dev. It would had taken me ages to fix third party 3D models where each dev have his theory of how UVs work, and most of the time, like never, they will not conform to game engines specifications. I can now throw at unreal a model and enable nanite and it works.
One of the main points is, it only renders what you want. If there is no inside, nothing gets rendered.
You can open the console and type freezerendering, and then moving around in the scene you can see what was being rendered at the time. But yea of course visibility culling is a thing 
As far as I know though, frustum culling works on a per object basis, and not per face, so if even 1 pixel of a huge object is on the screen, the whole object will be rendered, but that’s not something you can avoid