KSP2 System Requirements

[kerbiloid]

30 minutes ago, Vegatoxi said:

But they cant change all the math by the click. And wont do that just because simpler would be made a new game.

And the next project needs that work time.

More likely, the system requirements will be growing up on adding the listed features. So, the "80%" should not relax.

7 minutes ago, Gkirmathal said:

If that meant: KSP2 at low compares to highest graphics of stock KSP1 + following mod (mods configs modified for 6Gb vram GPU's): Parralex + Scatterer + EVE beta Volumetric Clouds &+ Waterfall exhaust shaders. Then you can run KSP2 graphics @minimum.

So, then it adds these features with no alternative, like hard-coded. While they are provided by system tools which can be disabled , like in KSP-1.

But in KSP-1 you can disable them by not installing the mods, in KSP-2 you probably can't.

Edited February 19, 2023 by kerbiloid

[Guest]

If you people come up with a set of craft files and interesting test I'll be more than happy to execute them on the Steamdeck.

[intelliCom]

Just now, purpleivan said:

 

 

You posted twice, I think you should just edit the first post.

But it's good, exactly what I suggested further above. Right here:

[purpleivan]

2 minutes ago, intelliCom said:

You posted twice, I think you should just edit the first post.

But it's good, exactly what I suggested further above. Right here:

Argh... a while since I edited a post here and the forum editing behaviour seems to have changed since I last did.

Fixed that now... thanks!

I was guilty of quickly skimming through a new and interesting thread and missed your post. Mine does indeed cover a lot of the same ground as yours.

One of the main points I wanted to make in my previous post, was that the camera view when FPS times are noted is going to be critical for the usefulness those FPS numbers, hence my including that in the my test descriptions. That's going to be a key factor in determining FPS in scenarios that include visual FX , as well as the time since effects started for things like enging plume FX, as these build up over time, affecting FPS accordingly.

[Gkirmathal]

26 minutes ago, kerbiloid said:

So, then it adds these features with no alternative, like hard-coded. While they are provided by system tools which can be disabled , like in KSP-1.

It would be bad if you could not disable such visual settings completely IMO. Like fully disable clouds shaders/lighting shaders/particle shaders etc for example.

Question is context.
What we've been shown, how does that compare in terms of GPU performance vs a modded KSP1 with visual mods that add these similar features? Features that seem to be 'stock' in KSP2.

What we have been shown in leaks, was that on minimum visual settings, or was that at max visual settings and on what hw?

So yeah the context is where they failed when announcing the hw requirements and where they can still "repair" all the fuss before e.a. date by giving the community such comparision as context.

 

Edited February 19, 2023 by Gkirmathal

[intelliCom]

Just now, purpleivan said:

One of the main points I wanted to make in my previous post, was that the camera view when FPS times are noted is going to be critical for the usefulness those FPS numbers, hence my including that in the my test descriptions. That's going to be a key factor in determining FPS in scenarios that include visual FX , as well as the time since effects started for things like enging plume FX, as these build up over time, affecting FPS accordingly.

I was considering having "aim the camera straight down" as part of the test. However, that wouldn't resemble the usual gameplay that a player has.
That being said, if  @MARL_Mk1 is up to discussing how to prepare some benchmarking data collection (Google Forms, test cases, craft files, etc.), then him, you and I could discuss the test cases more heavily.

Also the VFX point is why I had the "Engine Cluster from Hell" test. Every single engine brings out every single kind of engine plume. If there are huge performance dips when there shouldn't be, one of the engines could be causing performance issues for some reason.  So it doubles as bug testing.

[LoSBoL]

31 minutes ago, kerbiloid said:

And the next project needs that work time.

More likely, the system requirements will be growing up on adding the listed features. So, the "80%" should not relax.

Just an feeling here that the minimum specs wont go up further as EA development progresses and that the now given minimum specs are prognosed for the full game with colonies and interstellar travel. 

Why do I have that feeling? Bumping minimum specs during development will continue to have a row across the community whenever they do. 

I think the game will be getting more and more demanding when more features are added, and there wont be to many performance issues initially at the start when you have a 150 part launch vehicle on the launch pad when going below minimum specs.

We'll know in a week how good performance will be on the pad with a 4GB GTX970 at 1080p, someone is bound to try. 

 

Edited February 19, 2023 by LoSBoL

[Guest]

10 hours ago, Jacke said:

there's a high chance this is the same as Odyssey but worse.

I've seen ED Odyssey pop up a few times around here.

The main problem with Odyssey for a lot of people wasn't the performance, but the fact that, for some reason, they decided to add a crappy FPS Planetside 2 wannabe in my scifi space flight sim  game, as the first major update in forever, dropping VR along the way.

It's like Eurotruck simulator biggest update being the ability to walk out of your veichle to engage in bar fights along the road. It doesn't make any sense in the context of the genre you're playing.

[purpleivan]

Just now, intelliCom said:

I was considering having "aim the camera straight down" as part of the test. However, that wouldn't resemble the usual gameplay that a player has.
That being said, if  @MARL_Mk1 is up to discussing how to prepare some benchmarking data collection (Google Forms, test cases, craft files, etc.), then him, you and I could discuss the test cases more heavily.

Also the VFX point is why I had the "Engine Cluster from Hell" test. Every single engine brings out every single kind of engine plume. If there are huge performance dips when there shouldn't be, one of the engines could be causing performance issues for some reason.  So it doubles as bug testing.

I think having  a set of games saves with the desired angle/distance already set up, with "don't move that camera!" as part of the instructions would be the way to go. That would remove small differences in framing of the scene and FX having an effect on the results.

No idea of how straightforward sharing gamesaves is in KSP2 though.

I'd be up for providing my own results for whatever tests are agreed on, once  the game is released, so I have access to it.

BTW... when I started writing up my test suggestions, I wasn't considering using game saves as a way of standardisng the camera view. Therefore my suggestions were a bit on the lightweight side (so no "Engine Cluster's from hell"), but if we make use of saves, then we'd get the range of tests that we'd want and reliable results.

[intelliCom]

1 minute ago, purpleivan said:

BTW... when I started writing up my test suggestions, I wasn't considering using game saves as a way of standardisng the camera view. Therefore my suggestions were a bit on the lightweight side (so no "Engine Cluster's from hell"), but if we make use of saves, then we'd get the range of tests that we'd want and reliable results.

Do saves require specific craft files to be downloaded, or do the crafts just exist with all parts there? We're still unsure of how they work in KSP2.

I'm certain sharing craft files is still a thing, hence why I included that in my test cases.

[Infinite deaths]

I hope my RTX 4090 can handle this. 

 

No mention of 4K resolution? 

[purpleivan]

Hopefully gamesaves do include the parts for the craft in the saves, the same as in KSP1, but I don't have any info on that.

Anyone else know if this is how things work in KSP2?

[intelliCom]

8 minutes ago, purpleivan said:

Anyone else know if this is how things work in KSP2?

No one knows because it's not out yet lmao

[KerbolExplorer]

34 minutes ago, Infinite deaths said:

I hope my RTX 4090 can handle this. 

 

No mention of 4K resolution? 

Looking at the game requiring a 3080 for 1440p, I could see a 4090 struggle at 4k

[panzer1b]

Well good thing (after checking benchmatrk comparisons), my 4yr old lappy should handle the CPU reqs just fine (its like 20-50% better then the i5-6400).

What i'm hoping for is that the flashy grafix (like reflections, parallax, ect) can be 100% FULLY disabled as then it just MIGHT run on my POS GPU (1050 TI).  Afterall, KSP1 runs within reason (gets painful above 1K parts but it will still RUN) with parallax 2.0 (scatters only, no actual parallax rendering as its not needed imo with the scatters enabled)+scatterer+Eve (including the recent true volumetrics)+TURD+waterfall and looks no worse then KSP2 in my opinion (never liked the overly shiny looks of KSP2 craft, hopefully they allow us to edit the specularity of the paintjobs to make them look moe like KSP1 does at stock texture settings).

Heck, solong as they bothered to optimize in the slightest degree it should still in theory run even on my garbage GPU.  Cpu, specifically odeling high part count interactions is where KSP1 and likely KSP2 is gonna be bottlenecked since i just dont see there being an efficient performance friendly way to model 1000s of part on part physics interactions ACCURATELY while not being a lagfest.

Now if only a new computer didnt cost as much where i live (thats the downside of exchange rates and living in Poland, a brand new reasonable gaming rig is the cost of 1-2 decent used cars), ten i might even upgrade, but given that my current 1 is still hanging on and it itself wasnt cheap i'm defenetely not upgrading just for 1 game : ).  When it shats the bed i'll likely get a rig that can easily handle KSP2 (as it makes no sense to skimp on computer capability given that any sane person buys a comp for at least 5 years of use), but lets hope i can at least run KSP2 with basically all the grafix reduced as it would be super cool to have such an old POS run a game this new : )

Edited February 19, 2023 by panzer1b

[Kerbart]

But a GTX 1070 Ti doesn’t seem that expensive, so it’s not so bad after all?

[JoeSchmuckatelli]

11 hours ago, MARL_Mk1 said:

Hi there. This is mainly a thought right now, but this idea could help most players in doubt on the search of the truth this upcoming Friday: "CAN MY COMPUTER RUN KSP 2 EA?"

The recent release of the Minimum and Recommended Requirements to run the game has set the community on fire. At the moment, and knowing that said requirements might be, at the very least, somewhat consiervative, all we know is that KSP2 could become one of the most demanding titles of this year (at least during the first few months of Early Access while further optimization work takes place).

For this reason, I thought about setting an organized thread on the 24th based on a simple template that users willing to contribute would use to describe their system specs and performance after using, say, 2 or 3 of the vessels that already come with the game in a couple of similar scenarios, essentially reproducing the same ingame scenarios on as many and diverse kinds of systems as possible. (I.e.: Launching the KerbalX from the launchpad to orbit, Landing X big craft on the Mün, Flying this big plane around the KSC, Crashing a big part count vessel at high speeds, etc.).

This could also prove to be of actual value to Intercept, since they could consult how certain aspects of the game's performance vary based on the hardware used in a more organized way than the average Steam Forum Thread user screaming things like "This ******* ** game **** *** wont work on my 970!!1 u scammers!".

The following is an example of said template with made up info. This template, and maybe a small commentary about the experience below it would be the only desired pieces of information in the comment, aimed exclusively to provide the performance perspective on a wide set of PC builds from as many users as humanly possible. This format would provide the reader with the exact info they are looking for without having to browse pages upon pages of people quoting eachother on irrelevant information for them.

________________________________________________________________________________________________________________________________________________________________

• System Specifications:
  
  -OS: Win 10 
  -CPU: i7-8700
  -GPU: RTX 2070
  -RAM: 16Gb DDR4 2133Mhz
   
• KSP2 Performance:
  
  -Graphic Preset: Medium (And maybe repeat this same point with different quality presets again, even with Custom variants if inclined to do so).
  
  -Scenario 1 (KerbalX to LKO): Mid 40fps during launch and ascent, 50-60fps on LKO
  -Scenario 2 (Big 40-50 part Plane around KSC): Stable 40-50fps, some freeze/stutters when crashing against the ground with many parts
  -Scenario 3 (Landing X big part count craft on Mün): Stable 40-60fps on Mun Orbit, some stuttering when closing in on the ground while it loads, smooth high 40s again when touching down.
  -Stress Test Scenario (Launch a 100 part vessel / Cheat said vessel to X Celestial Body's surface/ Cheat a really big Space Station to LKO): Blah Blah Blah same thing ^
   
• Further Comments: "RAM usage seemed to spike a lot after 20-30 minutes of doing the same thing. Looking at Kerbin while in orbit would decrease my FPS by X amount. Decreasing/disabling scatters seemed to improve FPS a lot, blah blah this blah blah that you get the drill."
   
• Addition Video could be provided if the user desired so to get a visual grasp of any issues or bugs under said specifications.

________________________________________________________________________________________________________________________________________________________________

Ideas?
Suggestions?
Snacks?

PS: If anyone at @Intercept Games is willing to adopt this idea in an official way, they'd be more than welcome to do so, since it would encourage many more players to share their experiences regarding performance, and I'm sure they'd be more than grateful of being provided said information voluntarily in an organized manner for the benefit of all: dev team and players.

If you are going to do this - you need to hammer into people the requirement of including their screen resolution.  Then you must separate results into the categories of 1080p, 1440p and 4k. 

Example: Knowing that player JS has a 3070 but is only getting 30fps with a 200 part-count craft is disingenuous to the average player unless you also know he's rocking a 32 inch 4k monitor. The average person only knows FPS as a number and does not think past that. 

It will be incredibly unfair to the Dev team for people to start shouting about how 'badly optimized' KSP2 is because players with 3k series GPUs are getting low frames, where the misinformed presumption is that affected everyone, equally.  However once you start scratching the surface and you discover that a 3070 at 1080p is fine - even optimal - the damage will already be done. 

The industry is in a transition state from the legacy of 24 inch monitors at 1080p to larger monitors at 1440p and 4k.  Most people don't recognize that. 

Please collate results by the number of pixels that the GPU is being asked to push! 

[Kerbin Launch Coalition]

36 minutes ago, KerbolExplorer said:

Looking at the game requiring a 3080 for 1440p, I could see a 4090 struggle at 4k

That's crackers, to suspect an RTX 4090 struggle with a game that looks like this one is bordering on comedy. The scary thing is, I think you're spot on given the system requirements.

[JoeSchmuckatelli]

Please report your KSP2 performance here ONLY FOR 1080P resolutions. 

Include:

Spoiler

GPU (you must specify mobile variant, if applicable) 

CPU

RAM 

Storage type (where game is) (NVMe, SSD, HDD) 

Please discuss loading times, FPS and other results here only if you play at 1080p resolution. 

[JoeSchmuckatelli]

Please report your KSP2 performance here ONLY FOR 1440p resolutions. 

Include:

Spoiler

GPU (you must specify mobile variant, if applicable) 

CPU

RAM 

Storage type (where game is) (NVMe, SSD, HDD) 

Please discuss loading times, FPS and other results here only if you play at 1440p resolution. 

[JoeSchmuckatelli]

Please report your KSP2 performance here ONLY FOR 4k resolutions. 

Include:

Spoiler

GPU (you must specify mobile variant, if applicable) 

CPU

RAM 

Storage type (where game is) (NVMe, SSD, HDD) 

Please discuss loading times, FPS and other results here only if you play at 4k resolution. 

[JoeSchmuckatelli]

Edit - I started the relevant threads. 

 

Folks participating... Please keep your results in the appropriate threads! 

 

 

[magnemoe]

10 hours ago, MARL_Mk1 said:

The idea has its roots on the fact that we still lack context behind those Requirements.

KSP is a game that could choke a i9 and 4090 if you added ENOUGH parts. Unlike traditional games, there is not a ceiling of lowest performance, but a "keep adding parts until it crashes" sort of environment. A true sandbox.

We know that Minimum Specs ask for an i5 6400, 12gb RAM and a 2060 for 1080p Low, but under which conditions? Under which scenarios, vessel part counts and locations was it concluded that said system was the minimum you should have for the game to... run?

Was it with a full KerbalX on Mun's surface at 60fps? Was it with an Interstellar ship on LKO in one of their dev builds with unreleased content trying to aim for what said specs would be in a few months? Was it just to load KSC with a Thumper, Mk1-3 and a Parachute at +30fps?

I know for a fact that many people will try KSP2 on way, WAY lower spec systems than the minimum that is shown officialy.

I can't think of any other way to test the real minimum/stable requirements than to get together and do a proper Community Vibe Check to that first Early Access build.

I like this. Although recording video might induce an inherent performance loss to an already low spec system

I'm pretty sure the requirements is for interstellar ships and colonies. 
If they set it for the stuff we seen in beta screenshots and videos, the KSC and kerbin itself would be the most demanding. 
They can test this to some degree now, its just building a real massive ship and port it to orbit. 

If they set it lower and it ended out that people buying the game could not do the endgame they would be in much more problems. Serious backlash and might even getting sued. 
 

[JoeSchmuckatelli]

Please report your KSP2 performance here ONLY FOR Widescreen, Multi-Monitor and non-standard resolutions. 

Include:

Spoiler

Pixel and screen dimensions / screen setup 

GPU (you must specify mobile variant, if applicable) 

CPU

RAM 

Storage type (where game is) (NVMe, SSD, HDD) 

Please discuss loading times, FPS and other results here only if you play at Widescreen / Multi-Monitor or other non-standard resolutions. 

Edited February 19, 2023 by JoeSchmuckatelli

[GoldForest]

8 minutes ago, JoeSchmuckatelli said:

Please report your KSP2 performance here ONLY FOR Widescreen, Multi-Monitor and non-standard resolutions. 

Include:

  Reveal hidden contents

Pixel and screen dimensions / screen setup 

GPU (you must specify mobile variant, if applicable) 

CPU

RAM 

Storage type (where game is) (NVMe, SSD, HDD) 

Please discuss loading times, FPS and other results here only if you play at Widescreen / Multi-Monitor or other non-standard resolutions. 

All screens today are wide screen, so, you're asking for all performance metrics