This is a copy of the Java documentation for libGDX. It is not finished being ported.

All of the ApplicationListener methods are called on the same thread. This thread is the rendering thread on which OpenGL calls can be made. For most games it is sufficient to implement both logic updates and rendering in the ApplicationListener.render() method, and on the rendering thread.

Any graphics operations directly involving OpenGL need to be executed on the rendering thread. Doing so on a different thread results in undefined behaviour. This is due to the OpenGL context only being active on the rendering thread. Making the context current on another thread has its problems on a lot of Android devices, hence it is unsupported.

To pass data to the rendering thread from another thread we recommend using Application.postRunnable(). This will run the code in the Runnable in the rendering thread in the next frame, before ApplicationListener.render() is called.

new Thread(new Runnable() {
   @Override
   public void run() {
      // do something important here, asynchronously to the rendering thread
      final Result result = createResult();
      // post a Runnable to the rendering thread that processes the result
      Gdx.app.postRunnable(new Runnable() {
         @Override
         public void run() {
            // process the result, e.g. add it to an Array<Result> field of the ApplicationListener.
            results.add(result);
         }
      });
   }
}).start();

Which SharpGDX classes are Thread-safe?

No class in SharpGDX is thread-safe unless explicitly marked as thread-safe in the class documentation!

Particularly, you should never perform multi-threaded operations on anything that is graphics or audio related, e.g. use scene2D components from multiple threads.

HTML5

JavaScript is inherently single-threaded. As such, one of the limitations of the HTML 5 backend is that threading is not possible. Web Workers might be an option in the future, however, data is passed via message passing between thread. Java uses different threading primitives and mechanisms, porting threading code to Web Workers will not be straight forward.