Install Steam
login
|
language
简体中文 (Simplified Chinese)
繁體中文 (Traditional Chinese)
日本語 (Japanese)
한국어 (Korean)
ไทย (Thai)
Български (Bulgarian)
Čeština (Czech)
Dansk (Danish)
Deutsch (German)
Español - España (Spanish - Spain)
Español - Latinoamérica (Spanish - Latin America)
Ελληνικά (Greek)
Français (French)
Italiano (Italian)
Bahasa Indonesia (Indonesian)
Magyar (Hungarian)
Nederlands (Dutch)
Norsk (Norwegian)
Polski (Polish)
Português (Portuguese - Portugal)
Português - Brasil (Portuguese - Brazil)
Română (Romanian)
Русский (Russian)
Suomi (Finnish)
Svenska (Swedish)
Türkçe (Turkish)
Tiếng Việt (Vietnamese)
Українська (Ukrainian)
Report a translation problem
The physics calulations in Unverse Sandbox 2 are very intensive. A lot of it depends on your CPU. However, I've heard that at some point, they will try to off load some of that to the GPU, which would speed up simulations greatly.
Hope this answers your question.
From the videos on this game I've seen, unless you detonate several supernova at once, you're not going to get a huge lag issue in this game.
Unless your computer is really old.
I suppose 1000+ planets near each other would cause significant lag too.
But from all I've seen, this game is not going to cause any issues on any modern computer, even with it's physics calculations.
It's only when Quantum physics get calculated that you'll need a better computer. :)
Mind you, that probably won't happen until Quantum computers are publicily made available in 15 years. :)
As others have remarked, the calculations will generally utilize your cores very well, and scale with the computing power you have.
Currently the calculations are running "managed mode" .net which is, after a complete nbody physics rewrite, quite performant, but since you can essentially keep dropping attracting bodies in, every system can eventually get into trouble.
Therefore a physics computation implementation has additionally been done in c++ for better performance, and soon we hope to add back a GPU computation mode again.
In a previous version we were doing the gravity calculations on the GPU using OpenCL but for various architecture reasons, that has been taken out for the time being.
ie, on a Core i7 @ 4GHz, 16 gigs of ram and dual 980s maybe the threshold would be 1000x accurately, but on an old Core Duo, it would warn you about exceeding say 50x.
I don't mean like a hard limit on how fast you could speed things up. Just like a marker indicating that you're golden up to this speed.
The game doesn't seem to lag unless you're blowing up multiple supernovas, or spawn 100s of objects close to each other.
Some of the speed settings used I've seen run into a whole lot more than 1000x times normal speed.
Some in game simulations run on a time scale of several hundred thousand years a second.
I'm just hoping this game can eventually support Tectonic drift.
Now we use so called adaptive integration, which essentially mean that you can set a threshold of acceptable error per step, and the simulation will respect that, regardless og how fast you try to make time.
Since we cannot do magic (a lot) if you have a complex simulation, a low error threshold and a slow computer, the simulation speed will drop. When it does, the little time speed indicator turns red, telling you that you may have requested the simulation to run at a thousand years per second, but you cannot get it. Hovering over that will tell you how fast you are in fact able to go, given the current simulation, computing power and error threshold.
Cheers to Thomas (Greenleaf) and the rest of the team for making the physics and CPU utilization work so much better than in the original Universe Sandbox.