Vertex cache optimization

[Itun.itu]

I have done this all.
I have two meshes (one index buffer and one vertex buffer for each mesh). Before rendering I optimize them. Then in each frame I have a loop with two iterations in which there is glDrawElements call. I don`t have materials, textures and others. My scene is very simple.
Note: I use JOGL.

[Dark Photon]

Gonna need more details to help you, but given that you said Java and simple (suggesting small batches probably), I would suspect you might be CPU bound submitting batches or state changes.

If it's simple, convert to C for testing, and see if any perf diff. If not, you've got something simple you can post for folks to review/try.