Including Dx 12 support

I would like to suggest to include newest technologie in the game and start supporting the latest Dx version.

The reason is, that one of my parks got to a stage where it's getting more or less unplayable, because the framerate won't go farther up as about 20-25 fps, regardless of graphic options and resolution settings.
I have been told, the reason lies in Dx 11 restrictions, that prevent the game from using the full power of the computer.

Since i don't think, that i'm the only one with this problem and the group of users being able to use Dx12 should be large (Windows 10, as the only Windows OS supporting Dx12, is used by about a third of all computer users), i think it would be a good step to further improve the game and make it able to handle larger amounts of guests in the park.
 
Implementing DX12 doesnt magically give you better performance. They would have to rebuild their engine for proper DX12 support.
 
The reason is, that one of my parks got to a stage where it's getting more or less unplayable, because the framerate won't go farther up as about 20-25 fps, regardless of graphic options and resolution settings..

I don't want to sound nitpicky here, but 20-25 Frames are far from unplayable. It's not like in a shooter where you are relying on fast reflexes. But yes, full Parks are getting laggy when a certain number of Guests is being reached. It is not a graphics card Issue at that point.
 
Last edited:
Well to me it is unplayable, since i do notice the game being not fluent, which especially sucks, when you want to "ride" a coaster you just build yourself. Also everything you do takes longer. Not to mention that it somewhat takes the fun out of an otherwise outstanding game.

And that it will take time to proper implement such a support goes without saying. But after all, working on the game is what the developers are payed for and so it seems to make sense to suggest to try and improve the games performance with larger parks. After all the went for making it possible to build them, so it would be logical to make the game able to run fluent with them.
 
I play on a low-end laptop with specs that barely meet the minimum requirements, yet surprisingly I have been able to play the game with parks up to about 6,000 guests. Obviously my frame rate is bad, and it does make the game horribly annoying to play when its slow and buttons are not responding. But you guys have to realize that its a good thing this game is laggy, because it means that it will get better in time. RCT3 was the same way, give it a couple years when we have faster CPUs and the game will run smooth as butter. The bigger concern is how future add-ons will effect performance. If PC ran smooth today, then it would mean that its not pushing the limits of our CPUs. Like others have said, its not the same thing as a multiplayer game
 
Last edited:
That's why I said in certain situations. For the park that was used to test this, the optimizations caused the fps to increase from around 3 to around 60.
 
I play on a low-end laptop with specs that barely meet the minimum requirements, yet surprisingly I have been able to play the game with parks up to about 6,000 guests. Obviously my frame rate is bad, and it does make the game horribly annoying to play when its slow and buttons are not responding. But you guys have to realize that its a good thing this game is laggy, because it means that it will get better in time. RCT3 was the same way, give it a couple years when we have faster CPUs and the game will run smooth as butter. The bigger concern is how future add-ons will effect performance. If PC ran smooth today, then it would mean that its not pushing the limits of our CPUs. Like others have said, its not the same thing as a multiplayer game

My CPU is not pushed to it's limit by the game. The core problem is, that it is impossible to increase the performance at a certain point (seems to be at around 6.000 guest, since i'm having about 6.500). As said above by Crowdpleaser, at that point the limiter is not the CPU or GPU (thus meaning that future and better CPU and GPU won't be able to increase the performance of the game with large parks), but the game itself.
 
it is impossible to increase the performance at a certain point (seems to be at around 6.000 guest, since i'm having about 6.500)

People have successfully run parks with 15,000+ guests

If you cant get a decent frame rate at 6,500 guests you must not have a top of the line CPU... it might help if you posted your specs since your talking about performance issues... like I said I have a low end laptop so its quite surprising I can get up to 6,000 guests but the game is capable of going much much higher
 
Last edited:
If the guys with 15k+ guests can run it fluently i'm sure as hell interested in how they made it work.

I just had a look into it, since i never looked at the numbers and CPU is at about 75% on average running the park, with peaks at 85%. So it's save to say, the performance is not being limited there. And since turning down graphics quality and/or resolution doesn't really net anything too, it seems clear, that the limiting factor is the game.
Doesn't make it a bad game, i find it to be the best park simulator since RCT3, but i would find it nice, if they worked on (and the forum post of Chems shows they do) increasing performance of the engine when parks get bigger.
 
I think 15k peeps might be exaggerated. But for lots of people low fps will not be a problem since this game doesn't really require fast response times. On the other hand, it is understandable people don't like low fps, since it's hard to get used to if you always play in glorious 144 fps. [tongue] 4-6K guests seems to be around the sweetspot for most people.

But guests aren't the only thing in this game that affect performance. The game has to render a lot of high detailed objects. Since the game uses Dx11, the draw calls for the GPU are handled by just one core, so eventually this will be your bottleneck. Dx12 would allow to use multiple cores for draw calls, which might give some performance improvements. But again, you can't just slap Dx12 onto software to make it run better unfortunately (not that you are saying this btw). It would require a complete rebuild to get the most out of it, else it is just a lot of extra work that might give some slight performance improvements, but could also cause the opposite in certain situations.
 
I think 15k peeps might be exaggerated. But for lots of people low fps will not be a problem since this game doesn't really require fast response times. On the other hand, it is understandable people don't like low fps, since it's hard to get used to if you always play in glorious 144 fps. [tongue] 4-6K guests seems to be around the sweetspot for most people.

But guests aren't the only thing in this game that affect performance. The game has to render a lot of high detailed objects. Since the game uses Dx11, the draw calls for the GPU are handled by just one core, so eventually this will be your bottleneck. Dx12 would allow to use multiple cores for draw calls, which might give some performance improvements. But again, you can't just slap Dx12 onto software to make it run better unfortunately (not that you are saying this btw). It would require a complete rebuild to get the most out of it, else it is just a lot of extra work that might give some slight performance improvements, but could also cause the opposite in certain situations.

I'm not really into the whole technical aspect of this but how many cores does a modern gpu have? Since if it can only use 1 core but it has 4 or even 8 cores for example, wouldn't you get 4 or 8x the drawcall bandwidth and thus 4 or 8x the performance? If you look at it like that it does seem worth it but I dont know if it will work like that and why it would/wouldn't
 
i wouldn't expect any changes on supported direct x versions before the release of Planet Coaster 2. for the actual version it makes no sense at all (cost-benefit calculation)
 
before the release of Planet Coaster 2

Thats going to be a looooong wait lol, by the time Frontier finishes PC they will (hopefully) release Planet Safari, and both games will probably have tons of add-on DLC (which I'm hoping are compatible with each other) so it would seem like it will be very very difficult to make a sequel to PC anytime soon that actually surpasses it. When they release PC2 if its not compatible with the workshop from PC1 then it will be at a huge disadvantage so, I just don't see a PC2 happening at least not for a few years :)
 
When they release PC2 if its not compatible with the workshop from PC1 then it will be at a huge disadvantage so, I just don't see a PC2 happening at least not for a few years :)

that's the bummer i forgot to mention ;)

but my hope is, with all the experience Frontier made with the actual version, it's ups and downs, they won't have to start from the scratch and may have a better focus on usability, not just on performance alone. so it shouldn't take years to develop once they've started (or released).

at least they shouldn't again misuse future customers as game-testers to gain feedback about what lacks within the gameplay :D
 
Last edited:
at least they shouldn't again misuse future customers as game-testers to gain feedback about what lacks within the gameplay :D
One would hope, but I don't know... some people like that crap

Just an idea I was wondering about... I wonder how well PC would have sold if it launched completely free, but each update cost $10 a piece (not saying this is a better idea, just an idea I had that I thought was interesting)
 
Last edited:
I'm not really into the whole technical aspect of this but how many cores does a modern gpu have? Since if it can only use 1 core but it has 4 or even 8 cores for example, wouldn't you get 4 or 8x the drawcall bandwidth and thus 4 or 8x the performance? If you look at it like that it does seem worth it but I dont know if it will work like that and why it would/wouldn't

I think you mean CPU instead of GPU? GPUs can have 1000s of cores, which are needed to calculate all the meshes/shaders/etc.

It seems logical that if you increase the usable cores from 1 to 4 that the performance would increase 4 times, but unfortunately this is not the case. Let's keep it simple, we have a program that has 4 functions (A, B, C, D) for the CPU, each core of our 4-core CPU can handle 1 function at a time. So we could give each core one function and everything will be done at the same time. But what if function C needs the results from A and B, and D needs the result of C? Now we cannot calculate them at the same time anymore.
 
I think you mean CPU instead of GPU? GPUs can have 1000s of cores, which are needed to calculate all the meshes/shaders/etc.

It seems logical that if you increase the usable cores from 1 to 4 that the performance would increase 4 times, but unfortunately this is not the case. Let's keep it simple, we have a program that has 4 functions (A, B, C, D) for the CPU, each core of our 4-core CPU can handle 1 function at a time. So we could give each core one function and everything will be done at the same time. But what if function C needs the results from A and B, and D needs the result of C? Now we cannot calculate them at the same time anymore.

Oh I thought he said GPU, but yeah GPU and CPU probably also can only communicate with themselves on certain levels. I'm just trying to brainstorm here. For me the current performance issues when making halffull or full parks just keep me from really enjoying it in the long run that's why for me personally dx12 would be worth it. I'd upgrade from win7 to win10 just for that. But in the other topic there was a post so there might be hope that it will run better even on dx11, we'll gonna have to wait to find out
 
Back
Top Bottom