Hello,
A lot of the RealityKit APIs (Ex. LowLevelMesh
, LowLevelTexture
, etc.) are marked with MainActor
so they needed to be accessed on the main thread.
This creates issues when we need to perform expensive GPU related operations since now we need to perform those on the main thread. This results in bottlenecks and hangs in our application. We would like to use a multi-threaded approach to solve these problems which is difficult to do here. We are constantly streaming data whether the app is just appearing or the user is interacting with our application so we need to be able to perform these operations on a separate thread.
Any advice on how to achieve this using RealityKit?
Thank you.