Gran Turismo's Future: "4K Resolution is Enough", But 240fps is the Target

Plus......who really cares about AI racing.....honestly I think that’s lowest on the totum pole as most sim racers race online not single player vs garbage AI.

4k, solid 60fps or higher, ray tracing, VR at 4K, dynamic weather, and improved tire model plus damage.
For me is important, i don´t like online races and something like that. I prefer a beautiful career mode like the past games but with decent IA.
No time or patience to play with others to the bumper cars.
 
I bet you would be surprised as to the numbers of gamers across different games and on even different platforms prefer racing a well coded AI than put up with the human wreckfest a lot of online racing turns out to be. Even prefer offline championships to online racing as well.
Actually even iracing is currently adding really adjustable AI racing within their service.
ACC has an very, very good offline adjustable AI to race against.
The AI in GTS is just too slow and likes to play the rubberband game but it would not take a lot to speed them up and make them more adjustable including aggression, make them more aggressive on holding/defending their position and to attempt to overtake once passed.

I find a good AI can give a good racing experience and in a lot of cases give you a much cleaner and more respectable racing experience than online. With the current state of the penalty system in GTS even with the poor AI in the game it is many times better than that daily sport mode race where you are constantly being punted off.

I guess you are one of those that think online makes a sim racer and offline is just someone playing games. At the end of the day both are just driving their pretend race car around a pretend track and really playing a game. lol!

The AI is so easy to outsmart on all of them though. It simply doesn't compare to racing against real humans.

Obviously I'm talking about online racing vs competent/clean drivers not 12yr old destruction derby players. That's what they should be focused on, improving the penalty system and including damage to nullify them.

If I'm not online I'm offline doing hot laps and learning the track. I have zero interest racing against AI when I can hop online and race against real opponents far superior to AI.

Honestly I couldn't give a toss about racing or playing online - I'll stick with the AI, thanks.

For me is important, i don´t like online races and something like that. I prefer a beautiful career mode like the past games but with decent IA.
No time or patience to play with others to the bumper cars.

Point is you guys are in the growing minority. Racing sims are overall abandoning the single player relative to the online play. Why waste resources on AI for the minority vs improve the game for the majority. That's why I said it's pretty low on the totem pole of priorities for game developers today.
 
Point is you guys are in the growing minority. Racing sims are overall abandoning the single player relative to the online play. Why waste resources on AI for the minority vs improve the game for the majority.

I think you are really fooling yourself thinking that those that prefer online is the majority as compared to those that prefer offline are the minority. I would feel pretty safe saying that you have those reversed.

There is a reason that GTS which was introduced as basically online focused only backtracked and started adding tons of offline content. Also it may be a clue that only a single digit percentage points of all GTS created accounts have played over 5 online sport mode races.

There is a reason ACC offers a full offline championship mode in their game.

Also Project Cars is also another heavy on the offline content.

But perhaps the one you should really pay a lot of attention to is that iracing which since its inception as a service has always been online only is NOW CURRENTLY ADDING AI offline racing within its service.

But I guess you know more than all of the studios combined and why they seem to feel that the offline content is important to include within the development of their games. I do not see games dropping offline content including AI but actually adding more of it to what they offer.
 
Point is you guys are in the growing minority. Racing sims are overall abandoning the single player relative to the online play. Why waste resources on AI for the minority vs improve the game for the majority. That's why I said it's pretty low on the totem pole of priorities for game developers today.
iRacing which was only online racing is now implementing AI.
 
A.I. is a mathematical model, it could even be run in an old console. More calculations per second doesn't mean better A.I. model.

No.

AI requires physics calculations as well in order to be interesting. In games like ACC or Rac
 
@Deadpool
If you want your racing game to sell, you better have SP content. Just look at the percentage of GTS online trophies, which show most people couldn't care less about MP and played the SP content in GTS.

GPU time, and the discussion is on display graphics, not any part of the game engine.
Framerate can be CPU dependent, too.

Really hoping we'll get dynamic time of day back at least for GT7. It's still astonishing in GT5/6
I am sure we'll see those element again, since PS5 support real time Ray Tracing.

The PS5 is supposedly selling at a loss when it hits the market. Which means whatever's going into it has got to be some really high end hardware...
It won't be really high end hardware. Rumours suggest we are looking at an 9-12TF GPU.
 
They are only now implementing after devoting resources where it mattered. Hence AI is a afterthought and low on the priority.

From doing some really unofficial searching the closest I could come up with is iracing has about 400,000 total memberships and roughly 100,000 active (currently active subscriptions) within the service.

Those total membership numbers have been total subscriptions acquired since the service was founded in 2004 and from my research the active account number of roughly 100,000 was from the latter part of 2019. Again these numbers are what I could come up with and I am not saying they are actually correct or documented.

I realize that iracing in the grand scheme of things is a very small operation and studio but I would be willing to bet that the addition of AI is NOT an afterthought or even low priority but rather the devs understanding that they were being limited in the growth of the membership by not offering an offline option.

Go on the official forums (must be a current member) and you will see the responses to the addition of the AI and even though very new the performance of this AI are very positive and there are a number of responses stating they renewed their stagnant memberships as a result. I also saw where supposedly that after announcing the AI that the new membership subscriptions has been better than it has in some time.

Again iracing could be devoting its resources to improving the quality of its car models or bringing its graphics level up to more current specs to its main competitions such as ACC but it has chosen to add in AI instead. Again these guys running this have extensive gaming studio backgrounds and they will not be adding something to the game as an afterthought but will add content or features they think it will help grow the product ad the service.

They realize moving forward they need to sell new memberships that will be buying content such as new tracks and cars to keep the doors open and the service viable.
This is the bad part of a subscription based service is that you constantly need the income stream from memberships to stay afloat. You do not have a brand new improved product to sell every couple of years that helps generate income.

Just my opinion but I think the last 6 years have shown they need to do something more than just the online if they want to grow the service to a size they can afford to improve the car models and graphics to become more mainstream rather than remain in the small niche market they are and have been in. Online will always be this services main focus but I think they see they need an additional offline focused membership to grow the service and the online.

But in order to become mainstream 10 year old graphics and cartoonish car models is not going to get you there either so my guess is they are counting on the new AI to produce the new members and increase cash flow to do the graphics and the cars. Again this is just based off my own personal opinions by what I observe currently happening.
 
I think you are really fooling yourself thinking that those that prefer online is the majority as compared to those that prefer offline are the minority. I would feel pretty safe saying that you have those reversed.

There is a reason that GTS which was introduced as basically online focused only backtracked and started adding tons of offline content. Also it may be a clue that only a single digit percentage points of all GTS created accounts have played over 5 online sport mode races.

There is a reason ACC offers a full offline championship mode in their game.

Also Project Cars is also another heavy on the offline content.

But perhaps the one you should really pay a lot of attention to is that iracing which since its inception as a service has always been online only is NOW CURRENTLY ADDING AI offline racing within its service.

But I guess you know more than all of the studios combined and why they seem to feel that the offline content is important to include within the development of their games. I do not see games dropping offline content including AI but actually adding more of it to what they offer.

Not fooling myself at all, just look at the trend for all games not just racing. They have all been focusing on online and putting offline as a afterthought. That's not a opinion that's a fact.

You're fooling yourself by looking at incomplete data and trying to derive a complete picture. Many people race online and don't do sport races, let's not forget about the online lobbies. We also have no metrics to look at regarding offline play and return players. Don't forget all of the players who buy GTS play a few times and never return.

The important metric would be to look at all active players. And for PD and other racing sims, they believe that their core audience are primarily moving to online play. The reason ACC and PCars offer offline content is the same reason most games still offer it, because there are enough players to warrant it. This doesn't mean they are the core audience, that can be seen based on where these games focus their resources...... and that's online racing. SP is not their priority.

You've fooled yourself into thinking that iracing adding offline content proves your point, when in reality it proves mine.

@Deadpool
If you want your racing game to sell, you better have SP content. Just look at the percentage of GTS online trophies, which show most people couldn't care less about MP and played the SP content in GTS.

This actually doesn't tell us anything and entirely relies on the assumption that people play online for trophies. I dont do SP at all, only MP and I could care less about trophies, which shows how that metric is meaningless.

From doing some really unofficial searching the closest I could come up with is iracing has about 400,000 total memberships and roughly 100,000 active (currently active subscriptions) within the service.

Those total membership numbers have been total subscriptions acquired since the service was founded in 2004 and from my research the active account number of roughly 100,000 was from the latter part of 2019. Again these numbers are what I could come up with and I am not saying they are actually correct or documented.

I realize that iracing in the grand scheme of things is a very small operation and studio but I would be willing to bet that the addition of AI is NOT an afterthought or even low priority but rather the devs understanding that they were being limited in the growth of the membership by not offering an offline option.

Go on the official forums (must be a current member) and you will see the responses to the addition of the AI and even though very new the performance of this AI are very positive and there are a number of responses stating they renewed their stagnant memberships as a result. I also saw where supposedly that after announcing the AI that the new membership subscriptions has been better than it has in some time.

Again iracing could be devoting its resources to improving the quality of its car models or bringing its graphics level up to more current specs to its main competitions such as ACC but it has chosen to add in AI instead. Again these guys running this have extensive gaming studio backgrounds and they will not be adding something to the game as an afterthought but will add content or features they think it will help grow the product ad the service.

They realize moving forward they need to sell new memberships that will be buying content such as new tracks and cars to keep the doors open and the service viable.
This is the bad part of a subscription based service is that you constantly need the income stream from memberships to stay afloat. You do not have a brand new improved product to sell every couple of years that helps generate income.

Just my opinion but I think the last 6 years have shown they need to do something more than just the online if they want to grow the service to a size they can afford to improve the car models and graphics to become more mainstream rather than remain in the small niche market they are and have been in. Online will always be this services main focus but I think they see they need an additional offline focused membership to grow the service and the online.

But in order to become mainstream 10 year old graphics and cartoonish car models is not going to get you there either so my guess is they are counting on the new AI to produce the new members and increase cash flow to do the graphics and the cars. Again this is just based off my own personal opinions by what I observe currently happening.

You are using anecdotal evidence which doesn't support your assumption. Cool some guys commented that they liked the addition of AI........ that's not a metric we can base anything off of. The fact that it has taken iRacing until now to even offer SP shows it's not the priority of their business model. iRacing also already has a great tire model/physics/online mode etc in place, this allows them the freedom to implement a SP mode without taking resources away from larger priorities. GTS on the other hand has a ways to go to catch up to iracing and AI/SP is low priority relative to the rest of the game. Is it nice to have? Sure, but there are far more important things to be dealt with first. The ranking system SR/Penalties are much higher priority.

From what I've read, the rationale for iRacing to add AI was also for use at venues where they showcase tech.

Regarding income stream, that same argument applies to GT or any game, subscription doesn't change that, and it arguably gives them better metrics to improve the game and see real time subscription increases. Their graphics are also pretty good, not cartoonish at all, not sure where you got that impression. I do think they are feeling the pinch from ACC and PCars and even GTS to a lesser degree. iRacing got caught resting on their laurels.

What is very very clear from all sectors of gaming, SP is lower priority when compared to MP. That doesn't mean there isn't a market for people who like SP, but that they are a smaller market.

Not all people though will play the game just to go race online. So I think it's better if it's a balance between the two.

Agreed. The point was that AI/SP are not as high of a priority as fixing the online, tire model, dynamic weather etc.

This is strictly opinion. But if we polled all of the active users, AI/SP would be low on the priority list. One day AI will be great, but at the moment it's meh, very easy to outsmart the AI.
 
Last edited:
Agreed. The point was that AI/SP are not as high of a priority as fixing the online, tire model, dynamic weather etc.

This is strictly opinion. But if we polled all of the active users, AI/SP would be low on the priority list. One day AI will be great, but at the moment it's meh, very easy to outsmart the AI.
Of course, we do know that online races have become common these days. But I hope it doesn't become a compelling reason to let the game turn into something mostly online-oriented for obvious reasons.
 
Of course, we do know that online races have become common these days. But I hope it doesn't become a compelling reason to let the game turn into something mostly online-oriented for obvious reasons.

I think once online is in a solid place, that will free them up to devote time to AI. The fact that it isn't a priority shows with how terrible it is across the board IMO.

Another stat I just noticed, 64% of all GTS players have completed less than 20% of the campaign. 35% of those are at 0%. Less than 20% of players have completed over 50% and only 13% over 80%.

The important stat we dont have access to, or at least that I cant find, is how many are playing in online lobbies. Private leagues etc.

To be fair, the campaign % isn't really a full story either. There could be players who like to just race against AI in random races now and then with no desire to do a campaign..... my guess is this is a low number, but it is a variable to consider.
 
Their graphics are also pretty good, not cartoonish at all, not sure where you got that impression.
Crank up ACC and then crank up iracing and compare the tracks, scenery, cars whatever you want to compare between the two and you will easily see where I got that impression. Graphically iracing looks to be at least 10 year old graphics and lighting as compared to ACC. Physics wise ACC holds its own pretty well also. They also have a full offline championship mode as well.

The big thing for me iracing currently has going for it is the physics are good and the tracks available that are available nowhere else although it does get pricey if you buy all the ones you want.

I will agree games such as FPS games are gravitating more to online and eliminating a lot of single player content but racing games although there are decent numbers online for some games I do not agree that or see the evidence that single player content is being removed. As far as AI is concerned it seems that racing games are constantly improving the performance and adding in traits such as adjustable aggression, performance which does not indicate a feature that is an afterthought or one that is going to go away anytime soon.

It is apparent that we disagree and that is not going to change about the popularity or number of players that prefer online or offline but as I have stated it seems to me that the focus is to add in offline events not eliminate them and this even goes for a service that has never in the past offered offline in any capacity at all.

To me what would be the golden grail is if ACC and iracing were to merge and take the best that both games offered with graphics and the online and offline and put it into one game that kept content pricing reasonable, now that would be an experience.
 
Rendering bloke here, for some more Mythbusters-style action!

It has something to do with it. 240fps costs a lot of CPU time, which would then be missing for advanced A.I.

GPU time, and the discussion is on display graphics, not any part of the game engine.

Actually CPU is used a fair bit for graphics, and use could increase significantly with higher framerates. While the brunt of the load would be taken by the GPU (simply by having to handle 4x the number of pixels drawn per second) a lot of work is still handled by CPU.

4K is more than enough. The human eye can't see more than 60 fps 1080i so there's really no reason for anything better. Sure, some people have very sensitive eyes but they're the exception. For most people 60fps 1080i is already good enough. I'll never understand the craze with 4K and 8K, etc. I've even heard talk of 16K already. Why?

Interlaced is super-easy to see problems with; watch some footage of a fighter plane or a jungle, things start to look like they've got sawtooth/fuzzy edges! In super slow-moving, super soft images it's ok-ish.


Beyond 4k (heck even beyond 1440p) the only really super-consistent way most people are able to tell is with text - smaller text 100% looks clearer at higher resolutions (effectively higher dpi, since screen size is usually similar). That's why phones and tablets have stupidly high resolution screens but TVs have been able to go so long at lower resolutions.

As for the rest of the graphics, are you getting 4x the improvement in visuals for the ~4x increase in load on the GPU and some extra on CPU? No, not really. There's better ways to blow the power available that are more noticeable.

A good test is this: what looks more realistic: a 4k gameplay capture, or a 480p youtube video (OF REAL FOOTAGE OF REAL THINGS FROM A CAMERA IRL) ? The only reason games aren't at 480p anymore is that we need some clarity and detail you can't get at 480p! Otherwise we'd dump every trick in the book we possibly could into super-realistic rendering, at a low res, then do the UI on top at full resolution.

Also CPU time, because physics would be calculated in a higher frame rate I reckon. You must admit, that it is very unlikely for PD to raise the framerate to 240fps, and leave the physics engine at 60, and/or not raise the refresh rate of the physics engine too. The poster made a point saying he would like the ressources rather spend on A.I. and other things, instead of 240 fps. He has the right to do so, as far as I understand it, as this is supposed to be some kind of...discussion and opinion sharing? Also, it is more and more common to calculate A.I. on the GPU, which gives the posters point "more weight".

No.

AI requires physics calculations as well in order to be interesting. In games like ACC or Rac

Not particularly. AI would sample the world-state at the start of it's cycle. That would be just the latest state the physics was at that particular moment in time. So position, velocity, angular velocity, upcoming waypoint, etc. Then it applies its determined action until the next AI thinking cycle (sorry I'm trying to translate this into Normal, it's not easy).

Unless you're talking about having more cars on track takes up physics time (due to it having to run suspension/tyrephysicsbody calcs for each car every physics update). Then, yes. But the 'AI-thinking' part, no, not really.


One of the big myths that gets around is that Everything has to be calculated in sync with Everything Else, and that's not only wrong, but the worst thing you can do.

You can do Important Physics (ie your car handling stuff) at say 400Hz, AI (thinking, decision making) at 50Hz, Secondary Physics (ie bollards cones and anything not important) at 30Hz (sidenote: I think GT5P/GT5 was doing it at 12Hz or something crazy low), Network stuff (sending/receiving position/info in multiplayer) at 32/64Hz, and Rendering at 60Hz and it'll all be pretty OK.

What happens if you DON'T do that is you get physics or AI or movement speed or UI or whatever that is 100% affected by framerate. You can ask any Dark Souls players how dumb of an idea that was. Or the Fallout players what happens to the UI at high framerates.


If anyone has technical questions, just message here and ask.
 
Last edited:
Don't want to be that guy, but seriously, is here even one person that believes GT7 will run at 4K 120fps with raytracing, ultra high quality textures, shadows and particles? High end computers equipped with a 2080ti can't do that and people expect a console with a regular 2080 equivalent to hit that? To do something like this they would either need to optimize the game at sadistic levels or run the game at a resolution less than 4K and upscale. I sincerely doubt GT7 will run at native 4K if they target 120 or even 240 fps. Of course we don't know how GT7 looks like, maybe the graphics aren't much more complex than GTS, if this is the case, then 4K 120fps seems doable. I would rather they stick to 60fps and a modest resolution such as 1440p and add dynamic weather, better AI and the likes, rather than going for native 4K and then having Kaz saying for once again the console isn't powerful enough. During the PS3 era he complained programming on cell was too complex, fair enough, during the PS4 era he complained the console wasn't up to the task, this was true as well, if he repeats the same story with the PS5, maybe he should be a bit more modest with the graphics and focus on the areas GTS lacked, most notably the AI. That's my 2 cents at least. Hopefully they will deliver a proper singleplayer mode as well, GT4 was almost perfect in that regard.
 
I think once online is in a solid place, that will free them up to devote time to AI. The fact that it isn't a priority shows with how terrible it is across the board IMO.

Another stat I just noticed, 64% of all GTS players have completed less than 20% of the campaign. 35% of those are at 0%. Less than 20% of players have completed over 50% and only 13% over 80%.

The important stat we dont have access to, or at least that I cant find, is how many are playing in online lobbies. Private leagues etc.

To be fair, the campaign % isn't really a full story either. There could be players who like to just race against AI in random races now and then with no desire to do a campaign..... my guess is this is a low number, but it is a variable to consider.

A recent article on here had a breakdown of online use from memory it was about 50/50 online vs offline 35% daily races and 15% fia races. I wouldnt read much into the campaign stats as its not very strong compared to earlier titles and doesnt reflect folks who do custom races in arcade mode. Everyone is different I find online racing stressful and I play games to relax and unwind each to their own. The good news is Kaz said they will be catering for both sets of fans on/offline. I really like the way they listen to the fanbase and have refined the game so much from our feedback, the handling has improved and the fan favourite tracks have appeared in spa and laguna seca. Bring on GT7!
 
Don't want to be that guy, but seriously, is here even one person that believes GT7 will run at 4K 120fps with raytracing, ultra high quality textures, shadows and particles? High end computers equipped with a 2080ti can't do that and people expect a console with a regular 2080 equivalent to hit that? To do something like this they would either need to optimize the game at sadistic levels or run the game at a resolution less than 4K and upscale. I sincerely doubt GT7 will run at native 4K if they target 120 or even 240 fps. Of course we don't know how GT7 looks like, maybe the graphics aren't much more complex than GTS, if this is the case, then 4K 120fps seems doable. I would rather they stick to 60fps and a modest resolution such as 1440p and add dynamic weather, better AI and the likes, rather than going for native 4K and then having Kaz saying for once again the console isn't powerful enough. During the PS3 era he complained programming on cell was too complex, fair enough, during the PS4 era he complained the console wasn't up to the task, this was true as well, if he repeats the same story with the PS5, maybe he should be a bit more modest with the graphics and focus on the areas GTS lacked, most notably the AI. That's my 2 cents at least. Hopefully they will deliver a proper singleplayer mode as well, GT4 was almost perfect in that regard.
That's my concern,i hope they don't sacrifice high graphical fidelty textures,dynamic tod and weather etc.,to reach the target of 120fps
 
That's my concern,i hope they don't sacrifice high graphical fidelty textures,dynamic tod and weather etc.,to reach the target of 120fps
If 4K 120fps is the target, they definitely will. The 2080 isn't a bad card, but even that can only get you so far. For example the 2080 can't maintain 60fps in Metro Exodus at extreme settings 4K with hairworks and RTX. Knowing AMD, their cards are notoriously unreliable, ranging from bad drivers and optimization to throttling and downclocking. Like I said, GT7 with GTS graphics is perfectly doable at 4K 60fps and 4K 120 fps might be achievable with the good optimization, but with dynamic weather, better AI, shadows, particles, lightning, textures and raytracing? Definitely unachievable.
 
I thought more about this from a 120-240hz TV or monitor ownership thing.

Who here owns a monitor or real world TV that does 120-144-240hz?

Its coming to a point where 144hz 24-32" monitors arent too expensive.

eg. a Samsung 27 1440p 144hz unit is around $300-$400 us here.

I would imagine this place wouldnt have that much owners of high end gaming lcds. Going further, is there a TV that supports 120hz real gaming mode? Not interpolated crap?

So here's where Sony can strike. Just say they release GT Sport II with 120-144-240hz support at these res... use it to spearhead Sony gaming TVs.

That's a strategy.

If it came to it, I would buy a Samsung 27" 144hz for GT Sport II.
 
For me is important, i don´t like online races and something like that. I prefer a beautiful career mode like the past games but with decent IA.
No time or patience to play with others to the bumper cars.
I totally agree a beautiful career mode like in the past games but with decent IA is very good :), this is where GT Sport fall down with the GT League and the category where the cars go is 👎.
I know GT Sport was mainly for online racing, but for players like you and me and many others that don't like online stuff like bumper car racing, this is where PD should have made a lot better career mode, and not the BS we got in the GT League which is below par.
 
No point increasing resolution if the human eye can't perceive it... I don't even understand the hype around 4K, because 1080 HD is more than adequate for all of us. However, the human eye sure to detect a faster framerate so yes, more power should be dedicated to that. But the bigger is issue is buying a TV that can actually support it.
 
No point increasing resolution if the human eye can't perceive it... I don't even understand the hype around 4K, because 1080 HD is more than adequate for all of us. However, the human eye sure to detect a faster framerate so yes, more power should be dedicated to that. But the bigger is issue is buying a TV that can actually support it.


I'm not sure if its confirmation bias but we have had reports where 4k in driving simulators does help in certain circumstances. People have said it helps in corners and it helps in looking down long straights.

In my own case we've had FHD for about 10yrs now and I know that I want for more resolution when I'm bombing down Mulsanne and you need to pick out LMP2 backmarkers in your LMP1 car at 330km/h...

I think we're not "there" yet.

4k/60 brings us closer.

There are studies that show that 1,080p is not even close to what clarity we see in real life.
 
No.

AI requires physics calculations as well in order to be interesting. In games like ACC or Rac

As in graphic area, the most important aspect is the talent programming, developper decisions and optimization.

You can create a very bad AI with 5x physics calculations than other.

That's why AI and physics are better in some old games than in some new games.
 
4K is more than enough. The human eye can't see more than 60 fps 1080i so there's really no reason for anything better.
I'd urge you to actually try out a high refresh rate monitor before claiming this. Even the jump from 60Hz to 75Hz is noticeable, going to 144Hz is huge. It might be diminishing returns higher than this, but humans absolutely can see the difference.

Same goes for resolution. A 27" 1440P monitor is significantly nicer to look at than a 1080P model of the same size.
 
Point is you guys are in the growing minority.

Wrong. 50-60% of PS4s do not have PS+. Single player games are some of the best selling games on PS4.

PD have already rightfully said that GT7 will be a mixture of past, present and future GT. Having a great SP campaign is what most people want.
 
Rendering bloke here, for some more Mythbusters-style action!





Actually CPU is used a fair bit for graphics, and use could increase significantly with higher framerates. While the brunt of the load would be taken by the GPU (simply by having to handle 4x the number of pixels drawn per second) a lot of work is still handled by CPU.



Interlaced is super-easy to see problems with; watch some footage of a fighter plane or a jungle, things start to look like they've got sawtooth/fuzzy edges! In super slow-moving, super soft images it's ok-ish.


Beyond 4k (heck even beyond 1440p) the only really super-consistent way most people are able to tell is with text - smaller text 100% looks clearer at higher resolutions (effectively higher dpi, since screen size is usually similar). That's why phones and tablets have stupidly high resolution screens but TVs have been able to go so long at lower resolutions.

As for the rest of the graphics, are you getting 4x the improvement in visuals for the ~4x increase in load on the GPU and some extra on CPU? No, not really. There's better ways to blow the power available that are more noticeable.

A good test is this: what looks more realistic: a 4k gameplay capture, or a 480p youtube video (OF REAL FOOTAGE OF REAL THINGS FROM A CAMERA IRL) ? The only reason games aren't at 480p anymore is that we need some clarity and detail you can't get at 480p! Otherwise we'd dump every trick in the book we possibly could into super-realistic rendering, at a low res, then do the UI on top at full resolution.





Not particularly. AI would sample the world-state at the start of it's cycle. That would be just the latest state the physics was at that particular moment in time. So position, velocity, angular velocity, upcoming waypoint, etc. Then it applies its determined action until the next AI thinking cycle (sorry I'm trying to translate this into Normal, it's not easy).

Unless you're talking about having more cars on track takes up physics time (due to it having to run suspension/tyrephysicsbody calcs for each car every physics update). Then, yes. But the 'AI-thinking' part, no, not really.


One of the big myths that gets around is that Everything has to be calculated in sync with Everything Else, and that's not only wrong, but the worst thing you can do.

You can do Important Physics (ie your car handling stuff) at say 400Hz, AI (thinking, decision making) at 50Hz, Secondary Physics (ie bollards cones and anything not important) at 30Hz (sidenote: I think GT5P/GT5 was doing it at 12Hz or something crazy low), Network stuff (sending/receiving position/info in multiplayer) at 32/64Hz, and Rendering at 60Hz and it'll all be pretty OK.

What happens if you DON'T do that is you get physics or AI or movement speed or UI or whatever that is 100% affected by framerate. You can ask any Dark Souls players how dumb of an idea that was. Or the Fallout players what happens to the UI at high framerates.


If anyone has technical questions, just message here and ask.

What is offloaded to the CPU specifically? Afaik GPUs are great at parallel processes and most linear algebra can be done in such a way. So why not have everything on the GPU?
 
Back