KSP2 Performance Update

[SSTO Crasher]

The dream is alive…

[Doc Shaftoe]

1 hour ago, Stephensan said:

please, in the name of everything kerbin, please add FSR and DLSS before release.. please i beg for the community.

I don't think the performance issues are going to be tied to rendering what we see on the screen. The slowdown in KSP has always been the under-the-hood calculations for physics and resources. I'm not sure FSR and DLSS will help.

[MechBFP]

1 hour ago, VlonaldKerman said:

Understood- but wouldn't this have been a priority for some time? I would think that in a rocket sim, optimizing calculations relating to fuel flow would provide downstream benefits and free up room for more complicated, advanced processes down the line. I'm an idiot when it comes to game dev so I may be completely wrong, but it seems weird that a system this integral is as un-optimized as it is, even in this early build of the game.

I also have the same question/concern. Perhaps some of the underlying code is used for other components like colonies and it is a side effect of that not yet being finished so it never got to the optimization stage. 
I can’t blame them for focusing on the long term goal but perhaps this is an instance where a bit of rework later is called for. 

[TheDartFrog]

hi! in SWDanny's videos on YouTube, I've noticed the same rocket lagging much more with struts than when it was all floppy, I wonder if perhaps the joints "trying harder" to stay together could be contributing to the lag?

I'm saying this because in another game (Scrap Mechanic), the devs recently added a 'Simple Physics' mode, which greatly weakens the joints (more floppy) but also greatly improving the framerate, so I wonder if perhaps a similar situation could be at play here

if that's the case, perhaps making strutted parts be hard-physics-parented could help with performance, and leaving physics joints only for rotors and whatnot

here are the two videos by SWDanny, with timestamps:
laggy but stable: https://youtu.be/5X2xAfZd7GU?t=109
smooth but floppy: https://youtu.be/ryph3269OjU?t=261

[Bej Kerman]

4 minutes ago, TheDartFrog said:

hi! in SWDanny's videos on YouTube, I've noticed the same rocket lagging much more with struts than when it was all floppy, I wonder if perhaps the joints "trying harder" to stay together could be contributing to the lag?

I'm saying this because in another game (Scrap Mechanic), the devs recently added a 'Simple Physics' mode, which greatly weakens the joints (more floppy) but also greatly improving the framerate, so I wonder if perhaps a similar situation could be at play here

if that's the case, perhaps making strutted parts be hard-physics-parented could help with performance, and leaving physics joints only for rotors and whatnot

here are the two videos by SWDanny, with timestamps:
laggy but stable: https://youtu.be/5X2xAfZd7GU?t=109
smooth but floppy: https://youtu.be/ryph3269OjU?t=261

lolwut?

I was under the impression floppiness contributed to a lot of the lag because of all the physics calcs. IIRC SimpleRockets2/Juno runs well enough on mobile because of the fact it doesn't do joint physics like KSP.

[Streetwind]

2 hours ago, AtomicTech said:

You might be in trouble:

https://gpu.userbenchmark.com/Compare/Nvidia-GTX-1660-vs-Nvidia-GTX-1070-Ti/4038vs3943

I'd like to quickly step in here and gently but firmly take you by the shoulders and steer you away from this website. It may look tempting as a quick and easily linkable online resource, but unfortunately it might very well be the least reputable site dedicated to gaming hardware anywhere on the planet. If you'd like to know why, articles and videos from tech enthusiasts tearing Userbenchmark apart are all over the net.

Please use literally any other resource to get your benchmark results. Anything but this.

[Zeek05]

1 hour ago, Jarin said:

I'd like lower, of course. But while the 1070Ti is a good card, it's also what... a seven year old one?

But yeah, like others are saying, so much will depend on physics bottlenecks, I guess.

GPUs 5 years ago still cost more than my gaming PC from 2015, I stuck with a RX 480 for years because I couldn’t afford any more parts. I saw a video of someone with a 4080 & 12 core i7 getting 20fps with a single rocket.. this is the sign of a very unoptimized game. Which is understandable, it’s early access, but it needs to move on FROM that, otherwise half the playerbase will not be able to run the game

[Strawberry]

2 hours ago, VlonaldKerman said:

Understood- but wouldn't this have been a priority for some time? I would think that in a rocket sim, optimizing calculations relating to fuel flow would provide downstream benefits and free up room for more complicated, advanced processes down the line. I'm an idiot when it comes to game dev so I may be completely wrong, but it seems weird that a system this integral is as un-optimized as it is, even in this early build of the game.

It's important to note that we know that the resource system in ksp2 is much higher scope compared to ksp1, according to the dev diary about it, a big part of the design is for colonies such as allowing resources to change while parts aren't in view. This is probably why the lag from these calculations is much worse, its much harder to optimize this with the greater scope of this.

[jimmymcgoochie]

Is there a website that does side by side comparisons of computer parts? All this talk of RTX9000s and such means diddly squat to me.

[Monger]

Yeah, you caused quite a ruckus with this slide. I get that you are tip-toeing around unrealistic expectations. Maybe it would have been better to publish this info after EA release, so others that are not bound like you can tell their story about performance. 

Tomorrow I will buy KSP 2, and will try on every single machine that I got to make it run, and let you know. 

[The Aziz]

4 hours ago, Grenartia said:

Out of curiosity, does an i5 4460 @ 3.2 GHz count as better, worse, or equivalent compared to the i5 6400 @ 2.7, for the purposes of these specs?

Probably worse. I used to have 4460, and even at full power it was slower than my newer CPU running at underclocked 2.8ghz. Generation matters, there's more to processing power than just cores and gigahertzes. Though with two generations behind it may not be that much worse. But if you aren't planning full upgrade to something new anytime soon, I'd suggest trying your luck at finding used i7 4790. Same generation so it will fit in your motherboard. Because gods that 4460 is old. And it used to struggle for me with KSP1 and few other CPU heavy games.

[SpaceJaM]

13 hours ago, sh1pman said:

An obvious solution to this is to use a pre-built reasonably sized rocket (e.g. Kerbal X) and test how well it performs on various PC hardware during launch. Then just publish the results.

Why can't we do it by ourselves?
Each one of us, when  trying the game, will test the Kerbal X rocket (or wathever we'll have in the game) and then publish in a common thread his pc specs and obtained framerate.

12 hours ago, Krzeszny said:

GTX 1070 Ti in minimum specs is still too high, but could be ok during EA. There's a lot of room for improvement.

The performance of large crafts could be significantly improved by adding the option to enable "simplified internal physics" which would disable joint flexing whenever the player wants. I've never really understood the point in flexing. In KSP 1 there's the mod Kerbal Joint Reinforcement Next and I was hoping for KSP 2 physics to be more rigid without struts, but that's just a personal thing.

I wanna play with the 8800 GTX from my GPU collection!
Or at least with the GTX 690.

[Anth]

11 hours ago, Bej Kerman said:

I was under the impression floppiness contributed to a lot of the lag because of all the physics calcs. IIRC SimpleRockets2/Juno runs well enough on mobile because of the fact it doesn't do joint physics like KSP.

The physics calculations are working all the time. They don't stop for a moment. Floppiness shouldn't decrease performance.

[Lisias]

20 hours ago, Klapaucius said:

I have a GTX 1060 at 6 GB so am slightly underspec with a 1070ti which has 8

I also have an I7 and 32 GB ram and a lot of disk space.  The PC was set up for video editing and was the best compromise on performance and affordabilty. It is also 3-4 years old now I think.

So, how much does exceeding in one area balance out (if at all) in other areas. My understanding is that KSP1 relied heavily on CPU processing but did not take advantage of multiple cores.  I'm assuming KSP2 is going to be better at using the whole system? Pardon my ignorance if I am making incorrect assumptions or asking the wrong questions.

I can't say anything about KSP2, but I know that is shares some things with KSP1 - as Unity.

Unity games usually relies heavily on single core performance - Unity, or at least the part of Unity that KSP1 heavily relies on, is heavily built [optimised] around a non threaded paradigm. As far as I know, the Unity's physics engine is not multi-thread capable - but this may be a weakness imposed by KSP1 itself, where all the Universe shares the same Frame of Reference at a given instant.

The weird thing about Unity, mainly 2019 at least, is that the more cores you have, the worst the performance gets - believe it or not. So, high frequencies, low core count CPUs are the best hosts for playing KSP1 and, probably, KSP2 too.

Memory is simple: the more, the best. I think that there's a point in which too much memory may choke the Garbage Collector (mainly due it's being hindered by the problem I described on the link above), but I didn't reached it yet.

Your main problem is VRAM. From KSP 1.4.3 towards 1.12.5, the main factor that screwed performance on my rig was the high VRAM consumption. Even lightly modded KSP1s are getting screwed lately when using 6GB VRAM GPUs due the excessive use of VRAM by KSP1 and, lately, PD-Launcher. I got very promising results on my really old MacPotato by downsampling the DDS textures on KSP 1.12.x - not to the point in which it performed so well as on 1.4.3, but still better (the new ground rendering introduced on 1.8.0 is also pretty heavier on my rig).

But, this is KSP1. Let's talk about what I hope to get on KSP2:

• Multiple Frames of Reference simultaneously.
  • With each main body having its own origin, everything under its Frame of Reference can be calculated on a separate physics engine instance, each one on its own thread, and so we can have some very perceptible improvements on performance on savegames with multiple vessels scattered on different stellar bodies.
  • Seweing the different Frames of Reference together will be pretty complex, but it will worth the pain IMHO
• Whatever Unity did screwed on 2019 that ended up getting worse performance as more cores you have, I really hope they had fixed it on Unity 2021.
  • This is just abhorrent. I have better performance and lower CPU temperatures by reducing the number of threads per CPU on KSP1 to a single one.
  • It's insane.
• Better Texturing Quality granularity than on KSP1
• Way better VRAM management than on KSP1
• Running on Windows 10.
  • Microsoft is heavily pushing newer hardware for Windows 11, no matter the performance specs of the rig

If KSP2 manages to tackle down at least 3 of the points above, we will have a better experience on KSP2 for sure - and if your rig can manage 1.12.5 at max settings, it should be able to run KSP2 in my humble opinion. Eventually, at least.

— — POST EDIT — — 

I rephrased a statement above. I think it's more… faithfull… to what I'm meaning. I hope.

Edited February 24, 2023 by Lisias
tyops as usulla...

[Elthy]

11 hours ago, jimmymcgoochie said:

Is there a website that does side by side comparisons of computer parts? All this talk of RTX9000s and such means diddly squat to me.

For GPUs i would recommend this:
https://www.3dcenter.org/artikel/fullhd-ultrahd-performance-ueberblick-2012-bis-2023

While its a german website the values should be easy to understand even if you dont speak it. They aggregate hardware reviews (with lots of benchmarks each) and calculate a percentage, 100% being some old GPU i cant remember. If you scroll down you will see values for older GPUs and a separate 4k Index, which is more relevant in high resolutions as the cpu can be a limiting factor in 1080p.

For comparing CPUs there is sadly no comparable score, the best i know if you want to compare recent hardware is this graphic: https://www.3dcenter.org/abbildung/performance-ueberblick-intel-raptor-lake

Again this is aggregated vom lots of reviews.

[The Aziz]

Perhaps one of the devs could answer.. How much will the game benefit from installing on fast SSD? I'm already planning to but if someone is on the fence on where to install, does it significantly speed up the loading process and all?

[rebel-1]

None of the AAA level games have such minimal graphics requests. 1070Ti for minimal settings? Will the graphics be better there than in Crysis or MSFS?

[GoldForest]

7 minutes ago, rebel-1 said:

None of the AAA level games have such minimal graphics requests. 1070Ti for minimal settings? Will the graphics be better there than in Crysis or MSFS?

Most AAA games go for legacy hardware support. KSP 2 is a modern game that is future proofing itself.

[The Aziz]

7 minutes ago, rebel-1 said:

Will the graphics be better there than in Crysis or MSFS?

You ever looked closely at the ground in MSFS?

[Arco123]

Scattering looks very stocky. The clouds are a definite improvement but look... Strange. Water is good though. 

I was hoping for a better physics engine but I don't see much. I think it will be improved eventually past KSP 1.  

With those high part count vessels coming in I think part welding is going to be a very necessary thing. 

 

 

[magnemoe]

Just now, rebel-1 said:

None of the AAA level games have such minimal graphics requests. 1070Ti for minimal settings? Will the graphics be better there than in Crysis or MSFS?

Most AAA games released today support PS 4, KSP 2 will not, the PS 5 gpu is equal to RTX 5700 XT or RTX  2070, yes its minimum specs but consoles get more out of hardware than pc and 1080 is standard 

[Minmus Taster]

T-2 minutes and counting

T-1 minute

[Bej Kerman]

4 hours ago, Anth12 said:

16 hours ago, Bej Kerman said:

I was under the impression floppiness contributed to a lot of the lag because of all the physics calcs. IIRC SimpleRockets2/Juno runs well enough on mobile because of the fact it doesn't do joint physics like KSP.

The physics calculations are working all the time. They don't stop for a moment. Floppiness shouldn't decrease performance.

That's what confused me about the idea of floppier parts causing more lag than ones that aren't floppy. However, when I said Juno runs well because it doesn't do joint physics, I really mean it doesn't do joint physics. It's not rigid parts running better than floppier runs like Frog said, rather the game just isn't doing the calculations at all. No matter how reckless you are, your rocket won't bend and parts do not do anything separately until the destruction of a part forces them apart.

[runner78]

3 hours ago, Lisias said:

Unity games usually relies heaving on single core performance - Unity, or at least the part of Unity that KSP1 heavily relies on, is heavily built around a non threaded paradigm. As far as I know, the Unity's physics engine is not multi-thread capable - but this may be a weakness imposed by KSP1 itself, where all the Universe shares the same Frame of Reference at a given instant.

I think you're not quite up to date, Unity itself uses a lot of multithreading on the engine side. Rendering uses its own thread, physics is actually multithreading so far i know. And with Unity's job system, a game developer can implement multithreading quite "easily" and is also used a lot internally by Unity.
But physics is generally not particularly well suited for multithreading and has some limitation.

[whaaw]

DRM Free Version this time