NOTICE: Before proceeding, be advised, this has been through many times on forums. But, read this regardless. Especially the Developer team. I provide more insight into the issue that should be on priority list and provide practical examples to support my claims. I by no means mean any disrespect towards rules or others, but I feel like this is the only way to get the attention this issue deserves. My apologies.
Yo. I'm assuming Update 21 will be released around June-July? That comes with a lot of interesting features that will surely keep us all busy. I've been waiting for the improved stellar model for ages now. It kinda sucks that as of now, when a star attempts to leave the Main Sequence, it instantly just blows up. This can be prevented by turning off the radius by composition but other issues arise. So that will surely be something to wait for, as is the planet grid model. I'm doing a lot of experimentation with creation of stars and different kind of binary systems. One I adjusted so at certain point in time you can see two suns rise from the horizon in an angle, looked cool, especially since one is a red dwarf. But I digress. One thing I like to do is study the effects of tidal heating and forces, and watch how an accretion disk forms. That is cool and all, but the amount of particles makes a huge problem. Serial processing used by modern CPUs isn't even close to being ideal for that job. That is why I'd like to ask the Dev team, even though you are very busy with new features (I'm sure), that you include parallel processing feature as an OPTION to the game, utilizing the raw horse power of a GPU. Thousands of CUDA cores or Stream processors simulating these things is gonna be a huge thing. In fact, I'm confident enough to say that you should add this to your PRIORITY LIST. It just happens that I have a few Maxwell Titans collecting dust on my shelf. This CUDA feature would allow me to expand my simulations to a next level. Galaxy creation failed due to CPU being maxed out. FPS was around 1-5. With the horsepower of those Titans, I could pull it off.
Also, I directed these tips, while not new suggestions, with practical examples and explanations, to the Dev team, but I'd also like to ask the entire US2 community about what kind of simulations you run, and how would parallel processing aid you with those simulations. Let's show the Devs together that CUDA acceleration is an important feature. As an extra example, even Kepler GPUs would demolish CPUs in US2 workloads. And for AMD, HD4000 series+ would do the same. So this is definitely something to consider, because for people on budget, you can find old GPUs for like 20 bucks on eBay, Craigslist or Amazon and for 20 bucks u get something that runs not only US2 better, but games in general. Devs, hear us out on this and make GPU acceleration happen. Doesn't have to be on your priority list, but this would most likely increase your sales because people with potatoes could also play this without lag on simple planetary simulations. So make this happen, as soon as possible!
Sorry for huge novel about a topic that's all over the forums, but I need to get your attention and provide good enough reasons for you to start working on this. I hope I succeeded.