Hello, im just wondering if it’s possible that the beamMP servers could be optimized for potato PC
The only thing i can recommend you to do, is turn on simplified collisions, turning graphics to extremely low, turning “foliage” or the plant stuff to zero, turning down tire marks, and only joining LOW pop servers. MAX 6 players. Every car should be accounted by CPU core. So if you have a quad core CPU, you will only be able to run 4 cars on the server. The only thing I can tell you is to get more ram that your motherboard can support, I have 32gb with an i7 8700k CPU which is 6 cores, and a 6700XT GPU, and I still run about 60 FPS with 8 players on normal graphics. Just need a better CPU and ram.
Correction: 1 core for the UI, 1 core for the graphics, 1 core for the audio, so at least 3 cores are reserved. This leaves you with 0 cores left on a 4 core CPU, because CORE0 is used by your system. You can see this in the Task Manager - main process, UI, graphics, audio and each subsequent vehicle has its own allocated core.
So the minimum recommended amount of cores would be 5 just to simulate 1 vehicle without any lag - every process gets its own full core.
Because if you spawn a vehicle when you are out of cores, the game allocates some threads from some cores that have “free” threads, be it from CORE0 (system), the core that is running your UI, audio (this is why it can lag sometimes), graphics, etc. So the physics simulation ends up being laggy, as well as everything else, including your system, because those “free” threads are obviously not a full core which is needed for 1 vehicle.
For the time being you can “unload” your cores by deleting other players - this way you will render them as spheres (blobs) - their last ping or location sent to the server. This is best done by disabling “Automatic queue” in the Multiplayer options, then in Vehicle Config > Debug > Remove all. To spawn a player back they will have to queue a different vehicle, then apply their queue manually.
Also if one to play on a Vulkan renderer, be wary of the GPU’s VRAM limits. 4 GB should be enough.
As for the RAM - 12 GB should be the minimum in this day and age, especially for BeamMP.
There are no FPS “configs” or “optimizers”, if you are lagging - that is your hardware’s limit. But if you DO meet the requirements and STILL lag, then there it is a software related problem - for example, check your drivers, close background tasks.
I run BeamMP on a MacBook with Intel on Windows with an eGPU, so can you!
You could even play on a Steam Deck, just look around!
So, it is not the BeamMP servers that are lagging, it is the way BeamNG is coded right now. The developer team have set their priorities - file compression and optimization are NOT one of them. It is taking a long time because there exist two kinds of code improvements: those that you add on top of the existing code and those that you have to rewrite from scratch. Some things in the game will never have features that one would want UNLESS the developers rewrite the code. Both ways inevitably lead to a dead end, a “futureless approach”. For example, ever wondered why the audio is simulating muffled audio ONLY around the driver’s camera (1x1m)? Because to implement audio-raycasting (real-time noise bleed when a door is opened, for instance) needs, once again, the audio code to be rewritten. This applies to anything, including but not limited to, optimization.
tl;dr it’s beamng, not beammp. optimization is not the priority. it is lagging because of your hardware.