Do you think KSP 2 will eventually end up more optimized?

[RandomGrape]

I think it probably will, but not for a while. Quite a lot of modern games have come out like this.

[The Aziz]

Yes.

[Pat20999]

Yes.

[pandaman]

Yes, I do.

[LoSBoL]

Yes, or else it will end up like with KSP1 with the same result. KSP1 got 'optimized' due to a decade of progression in PC hardware. When KSP1 got released the PC Hardware wasn't even ready for it to run 'optimised'.

[Alexoff]

With the release of the 8900ti KSP2 it will be normal to run on medium computers.

[Wheehaw Kerman]

Does anybody seriously think it won’t?

[Dakitess]

Yes it will. I expect it to be something like 2-300% more fluid as it is right now. Like going from 15 FPS to 45FPS. This is quite massive and it might take a whole 2 years from now.

But...

But I don't see how we will get some 1500% more performance which is actually required to step up (or even equalize) KSP1...

15 FPS at launch for 100 parts crafts with a 1000$ rig is not OK, 45 FPS will be "fine". But that's for a 100 parts.

KSP2 needs to allow for 1000 parts craft with somewhere like 20-30 FPS, with a decent machine. Let's say a 1000$ machine that would exist in 2 years.

This is a 1000 to 1500% performance gap, and nope, I definitely don't see how we will get this, even not factoring the fact that the game is supposed to enhance its graphics along the years, and add features and contents that won't simplify the optimization, definitely.

You'll say, I know, that 200 parts crafts are already totally fine, and 20 FPS is good for you. That's okay. But it's a massive step backward from KSP1, which I personnaly don't accept AT ALL, while the promise (yeah yeah source, blabla, etc) is some vastly more optimized performance, even / specifically for big crafts.

[DigitalEva]

It definitely will become more optimized as time goes on, no doubt about it.

[RocketRockington]

It'll get more optimized as bugs are fixed and performance is addressed - then less optimized as features are added and  graphics are improved and those cost performance.  Much of the current optimization has been about turning things off, after all.

  Given how many features are missing - especially multiplayer - and issues with the game's graphical state (terrible AA for instance)  - whether or not it will be more optimized than it currently is after(if) those features end up getting added is up in the air.

[Spicat]

18 minutes ago, RocketRockington said:

Much of the current optimization has been about turning things off, after all.

No, that's false

Edited June 12, 2023 by Spicat

[Davidian1024]

Yes, it will become more optimized.

[asmi]

On 6/11/2023 at 11:33 AM, Dakitess said:

KSP2 needs to allow for 1000 parts craft with somewhere like 20-30 FPS, with a decent machine. Let's say a 1000$ machine that would exist in 2 years.

Who said that? That was never promised, not it is in the roadmap, therefore it's very unlikely to ever happen. KSP1 still chokes on 1000 parts crafts today, despite huge progress in CPU and GPU horsepower, and I see no reason it will be any different with KSP2.

[Dakitess]

Same as with your other thread message.

Bring elements, otherwise you're only attacking the message and the author. I gave my feeling about it, with numbers to compare things accordingly. Do it as well if you don't think so, or express your opinion, [snip]

So, for you, what would be some key performance indicator to get for KSP2 ? 30 fps for 100 parts ? How is it supposed to scale with, say, 200, 500 and 1000 parts ? What's your take about it ? What do you think we will reasonably get ?

You don't want to think about 1000 parts crafts ? Okay, then tell us your performance expectations with your own criteria. This is what I did, feel free to disagree.

Edited June 13, 2023 by Vanamonde

[asmi]

On 6/13/2023 at 7:56 AM, Dakitess said:

Same as with your other thread message.

Bring elements, otherwise you're only attacking the message and the author. I gave my feeling about it, with numbers to compare things accordingly. Do it as well if you don't think so, or express your opinion [snip]

So, for you, what would be some key performance indicator to get for KSP2 ? 30 fps for 100 parts ? How is it supposed to scale with, say, 200, 500 and 1000 parts ? What's your take about it ? What do you think we will reasonably get ?

You don't want to think about 1000 parts crafts ? Okay, then tell us your performance expectations with your own criteria. This is what I did, feel free to disagree.

I would love it to support 1M parts vessels, but I know it's not gonna happen. And I doubt that 1K is going to become a reality either. Most of my crafts in KSP1 are on the smaller side because I like the challenge of achieving as much as possible with as little as possible parts (though I prefer using procedural parts which tend to reduce parts count compared to crafts created using regular parts). I think my largest craft was a space station with about 200 parts, but most "mission-oriented" crafts were/are <100 parts. If I get 60+ fps with such vessels, I'm happy.

But again, my point is that no hard numbers were ever given by devs, and so I try to set my expectations lower in order to avoid being disappointed.

Edited June 15, 2023 by Starhawk
Quoted post redacted by moderator

[Alexoff]

4 hours ago, asmi said:

Who said that? That was never promised, not it is in the roadmap, therefore it's very unlikely to ever happen. KSP1 still chokes on 1000 parts crafts today, despite huge progress in CPU and GPU horsepower, and I see no reason it will be any different with KSP2.

Here, 5:27

[Vanamonde]

Some comments removed. As always, please don't make things personal. 

[K33N]

The game needs a lot of work to meet the level of KSP1, and that is not taking into account 1000 part ships between stars and big bases on the planets.

Its going to take a lot of time and money, which I think is the crux of the question in this thread: how much money and time does KSP2 have? 

[asmi]

8 hours ago, Alexoff said:

Here, 5:27

I've watched it all, and nowhere had Nate said that 1000+ parts vessels are going to be possible at reasonable framerate. Infact he didn't call out any particular number, he just said "I hope that limit will be much-much higher than KSP1". Hope is not a promise, it's just that - hope. And for what it's worth, I hope that too. But software developer in me says that it's unlikely to happen any time soon, and I doubt this will be their focus at least until after few roadmap updates knocked out so that the game will be stable enough for such experiments. Right now even if you build a 1000 part vessel and somehow launched it into space in one piece, it will still probably tear itself apart at some point due to phantom forces, especially if you like save/load gameplay style.

3 hours ago, K33N said:

Its going to take a lot of time and money, which I think is the crux of the question in this thread: how much money and time does KSP2 have? 

That question is completely irrelevant to this thread, as no amount of money will make Cyperpank 2077 run on potatos with any quality to speak of.

[Guest]

On 6/10/2023 at 7:39 AM, RandomGrape said:

I think it probably will, but not for a while. Quite a lot of modern games have come out like this.

Certainly! I don't expect one "big bang" of an optimization, but a continuous series of incremental improvements, like we've already seen in the patches. Some of these will manifest as better performance, other as better visuals or new features.

[Alexoff]

On 6/13/2023 at 8:06 PM, asmi said:

I've watched it all, and nowhere had Nate said that 1000+ parts vessels are going to be possible at reasonable framerate. Infact he didn't call out any particular number, he just said "I hope that limit will be much-much higher than KSP1". Hope is not a promise

Yes, in fact, we were never promised anything concrete, only evasive phrases, hopes that the goals would be achieved. It seems the only thing that we were explicitly promised is the price of 50 bucks. [snip]

On 6/13/2023 at 8:06 PM, asmi said:

But software developer in me says that it's unlikely to happen any time soon,

I totally agree with this

On 6/13/2023 at 8:06 PM, asmi said:

Right now even if you build a 1000 part vessel and somehow launched it into space in one piece, it will still probably tear itself apart at some point due to phantom forces, especially if you like save/load gameplay style.

I will add that half of the parts of this craft will be struts. And I don't think it's good. Has anyone been able to fly to Eve and back from sea level yet?

Edited June 15, 2023 by Starhawk
Redacted by moderator

[Starhawk]

Some commentary has been removed.

As Vanamonde said above, please keep things from getting personal.
 

Thank you for your understanding,
Forum Moderation Team

[asmi]

On 6/14/2023 at 4:46 AM, Alexoff said:

Yes, in fact, we were never promised anything concrete, only evasive phrases, hopes that the goals would be achieved.

And that was very prudent of them, because whatever number they would've said would end up being wrong as it heavily depends not only on user's hardware, but software too, and you can get wildly different results on similar (on paper) systems.

[Alexoff]

6 hours ago, asmi said:

And that was very prudent of them, because whatever number they would've said would end up being wrong as it heavily depends not only on user's hardware, but software too, and you can get wildly different results on similar (on paper) systems.

Well, it's very prudent on their part not to promise anything specific, and on our part it's naive to expect that we will be able to see craft in 1000 parts with FPS above 30 on any computer. After all, no one actually promised anything!

[asmi]

16 hours ago, Alexoff said:

Well, it's very prudent on their part not to promise anything specific, and on our part it's naive to expect that we will be able to see craft in 1000 parts with FPS above 30 on any computer. After all, no one actually promised anything!

Exactly! We can hope for it, but expect - not.