As many of you might know server performance has always been an issue in the PU and that causes a variety of issues such as: unrensposive AI, vehicle/cargo clipping through elevators , etc.., part of the reason is that legacy CryEngine physics engine only uses 4 threads to process all physicalized entities and its not enough for a game with the scale of Star Citizen. At CitCon 2017 it was first mentioned (as far as I know) they were planning on using the batch system to improve performances of both the renderer and physics engine: https://www.youtube.com/watch?v=_-OoMySD66E&t=2148s We know the renderer transition to this system is currently in progress with the Vulkan refactor, but we haven't heard about the physics upgrade in a while and I'm amazed its rarely brought up by the community, cause it seems to be a crucial piece of tech needed to make SC an MMO on the same level as Vulkan, Server Meshing and iCache. CR talked about it extensively during 2 RTVs, first in 2018: https://www.youtube.com/watch?v=G42MQ1aVjlA&t=2612s and then in 2019: https://www.twitch.tv/videos/403077141?t=40m37s We haven't had any official update on the topic since then, the only recent reference I could find is in a monthly report from March 2020: https://robertsspaceindustries.com/comm-link/transmission/17543-Star-Citizen-Monthly-Report-March-2020 Which states the following: March saw Engineering working on physics threading and performance. They enabled up to 30 physics threads (defaulting to ¾ of logical cores available) The actual state of the refactor is unclear, it seems they have made some progress on it and perhaps have an internal dev branch that uses it, but its not quite ready for production, but then there is this clip of John Crewe stating the refactor was complete in 2019 and it was a blocker for docking, which guess what? We just got in the PTU. https://www.youtube.com/watch?v=_mawF0Gzy5o&t=2535s So it seems as if its already in game, but if thats the case why haven't we seen any improvements with server performance? CR seemed to be quite optimistic OCS + bind culling + physics refactor would allow the server to perform much better and even increase the player count up to 200 players. I'd really appreciate if we could get an update on the matter: Is it in game? If so why haven't we seen any tangible improvements? Is it cause game systems need to be converted to use it like it happened with Item 2.0/OCS or were the performance gains lower than initially anticipated?