Move some of the heavy combat math adaptively to the gpu.

[Therom]

So whenever we attack a single enemy vessel and we have an obscure amount of turrets on one ship, have the gpu take over that math to break down and generate loot efficiently and effectively. Dyson sphere program in the next update is going to do the same. the sort of math the gpu also can do will be moved to the gpu to help the cpu by lightening the workload and take over heavier math tasks that otherwise tanks the performance and framerate. so for me who has 100's of turrets on my ridiculous ship, have the gpu math out that targeted ship to boost performance.

It's like ubi did with AC odyssey to valhalla. Odyssey could even use up a 12 core thread ripper due to the amount of people and other physucs, but valhalla moved that over to the gpu thanks to DX12. Made any gpu tank, but the game is already graphically demanding. this game is not so much on graphics, so it has quite a reserve on gpu calculations.

[Kamo]

Awesome idea ! Turrets and aiming calculations are definitely a bottleneck, especially custom turrets for some odd reason. As soon as my fleet hits fifty or so custom turrets the game tanks 😖. With default turrets like on my TeaKon playthrough the game was still smooth with a fleet of 80-90 ships (it was a horde-like faction).
Large amounts of enemies (like ~60 entitydbg-spawned Xsotan in an asteroid field) also caused it too (while similarly large amount of ships on my side caused less lag).

Another thing that could be optimized is the UI, especially movement lines. Supply Factories lines on an entire Accelerator Factory supply chain with around sixty auto-freighters tanked my FPS on the Galaxy Map to 10-15 with huge input lag while it was 60 outside of action... except every minute when the actual supplying happened 🤭.

[AvorionCraft]

Are we playing the same game? Even running a 3600 this game still struggles.