• Hey, guest user. Hope you're enjoying NeoGAF! Have you considered registering for an account? Come join us and add your take to the daily discourse.

Digital Foundry: GTA V PS4 and Xbox One compared in new frame-rate stress test.

Did they have this freeway stack that the PS4 dropped frames on consistently vs XB1 in the video? They showed the firefight where XB1 dropped down to mid 20s for a good bit, but I never saw it on the PS4. Maybe I missed it.
 
Seriously with the sub 30fps stuff on current gen. Its really ridiculous. Shows you how underpowered both systems are, and particularly how crappy the CPUs are. And this is an upscaled remake to boot.

There is no way they make it past 5 years for this gen. The hardware can't support it.
 

Marlenus

Member
Not sure how they came to this conclusion when every segment minus two showed the X1 version dropping more frames even including high-speed chases.

IB4DKI1.jpg

That is the moment I was talking about above.

There is no real reason I can think of why the frame rate would be the same there. If it was CPU bound the Xbox should be ahead and if it was GPU bound the PS4 should be ahead. Either they still need to optimise slightly or there is another bottleneck in the system that is the same for both but I just cannot think what it is.

To be fair he did say that the Xbox One was ahead in high speed chases through busy junction sections so it is a qualified statement that is truthful.
 

Sorral

Member
Just curious but is there anything stopping Sony from upclocking the PS4'S CPU like MS has?

Yes, chances of bricking consoles if they weren't pre-made to handle that upclock like the PSP was. People touched on this earlier in the thread.

I see this CPU fud is working.
 

Teletraan1

Banned
Why did they add all that shit like extra traffic density and useless fluff if it couldn't hold 30fps. What a joke. Waiting for PC. Fuck my friends and controller preference. I have waited this long, ill wait longer.
 
Seriously with the sub 30fps stuff on current gen. Its really ridiculous. Shows you how underpowered both systems are, and particularly how crappy the CPUs are. And this is an upscaled remake to boot.

There is no way they make it past 5 years for this gen. The hardware can't support it.

It's really not that bad. The situations were that happens in GTA V are pretty rare compared to last-gen where it was a constant annoyance. And let's not even bring up the recent release of you-know-what by you-know-who as an example, that's just straight-up incompetence.
 
I would like to see some real-time monitoring of the cores in these consoles to see what the utilization is. I'm wondering how efficient these game engines are at multi threading.
 
This Digital Foundry stuff is great, but they need to stop pointing to the hardware exclusively when they're explaining things in these articles. Why should the PS4 version have the upperhand? Why should the overclocked CPU in the X1, make a difference?

The reality is that it's as much down to the code in the games, as it is to the power of the consoles. And these are ports of last gen games, I imagine it was optimised, but they weren't coded from the ground up for these machines. There will be a load of derelict code that while it doesn't break the program, still remains redundant on these machines.
 

JordanN

Banned
Neither can hold 30fps at 1080p? Sony is expecting the PS4 to last ten years? Sounds like a fun gen.

Developers have barely scratched the surface of these consoles. Having 16x more ram than last gen should tell you the best looking games wont show up till the end of the gen.
 

goonergaz

Member
I knew I wasn't going mad with the frame-drops...funny how initially they were saying the XBO version was litterally solid 30fps...just shows they are better off holding back until they can look into things properly
 

Kilau

Member
This Digital Foundry stuff is great, but they need to stop pointing to the hardware exclusively when they're explaining things in these articles. Why should the PS4 version have the upperhand? Why should the overclocked CPU in the X1, make a difference?

The reality is that it's as much down to the code in the games, as it is to the power of the consoles. And these are ports of last gen games, I imagine it was optimised, but they weren't coded from the ground up for these machines. There will be a load of derelict code that while it doesn't break the program, still remains redundant on these machines.

Yeah I really wish they would leave the editorial opinion out of the technical articles. Seems like a "perception becomes reality" thing is going on.

Hopefully for the full face off we will have a frame rate analysis breakdown per scene because this article is just very vague overall.
 
Neither can hold 30fps at 1080p? Sony is expecting the PS4 to last ten years? Sounds like a fun gen.

Last generation games are not indicative of the potential for current gen systems.

God of War remastered didn't run at a locked frame rate either. It had some deviation from 60.

If people actually took those dips literally as "PS3 being pushed too hard" then God of War III wouldn't have looked much different than the God of War remaster.
 

Hermii

Member
Developers have barely scratched the surface of these consoles. Having 16x more ram than last gen should tell you the best looking games wont show up till the end of the gen.
More like 10x. Current gen console os es reserve a ridiculous amount of ram
 

Teletraan1

Banned
I'd really love to know who started this myth. You can't "lock" a frame rate. You can cap it so it doesn't go over 30, but it can always be less than that. Especially in a game that isn't heavily scripted.

You can adjust your settings until you are comfortable with it never dropping below 30 and then cap it at 30. This is what I do on PC games for 30 or 60 frames, most recently AC:Unity. If I can get AC:U which is regarded as a broken piece of shit game to never go below a 33.3ms frame time on below minimum spec hardware then surely rockstar can do the same with their game. I do my diligence on PC, sad that console devs don't do the same, instead they add more cars and useless nonsense to the game and result in sub sub optimal frame rates so they can sell it to the double and triple dippers who only care about graphics.
 
This Digital Foundry stuff is great, but they need to stop pointing to the hardware exclusively when they're explaining things in these articles. Why should the PS4 version have the upperhand? Why should the overclocked CPU in the X1, make a difference?

The reality is that it's as much down to the code in the games, as it is to the power of the consoles. And these are ports of last gen games, I imagine it was optimised, but they weren't coded from the ground up for these machines. There will be a load of derelict code that while it doesn't break the program, still remains redundant on these machines.

Weren't they both ported from the same code(Xbox 360)? I don't think the code is the problem, unless it's just not well multi threaded.
 
Being CPU bound is going to be more common in games like this. The more AI, collision, and path finding going on at once the worse it gets. The GPU is probably not maxed out.
 

AgentP

Thinks mods influence posters politics. Promoted to QAnon Editor.
I would like to see some real-time monitoring of the cores in these consoles to see what the utilization is. I'm wondering how efficient these game engines are at multi threading.

This game exists on the PS3/360... not CPU monsters themselves, especially the PS3 (single PPU).

The bottlenecks probably resides elsewhere and come down to optimization, not something simple like CPU load. It's not like simple pathfinding for drivers and pedestrians need a lot of CPU.
 
Being CPU bound is going to be more common in games like this. The more AI, collision, and path finding going on at once the worse it gets. The GPU is probably not maxed out.

I'd say that the GPU is almost definitely not maxed out, we'll know for sure once the PC version hits.
 

On Demand

Banned
Looking in more detail at the video I do not think the advantage on Xbox One is CPU related. If you look at the video @ 3:43 the scene is almost the same on both consoles and so is the FPS. If this scene was CPU limited you would expect the Xbox One to be ahead slightly and if it was GPU limited you would expect the PS4 to be ahead. The fact that neither one is ahead suggests this is a case where a bit more optimisation is needed or there is a bottleneck somewhere else that is the same on both consoles.

It's really hilarious that everyone is going with this "CPU bound" thing now.
 

Alienous

Member
What's going on in this thread? The PS4's tablet CPU is a confirmed dud?

I don't want to be CPU locked so early in the generation. God dammit.
 

Marlenus

Member
Weren't they both ported from the same code(Xbox 360)? I don't think the code is a problem.

Even then it is a lot different. The IBM PPC in the 360 is a RISC based CPU using the Power architecture, the AMD Jaguar is a CISC based CPU using X86 architecture. They work in a fundamentally different way.
 

RoboPlato

I'd be in the dick
Seems like R* put in a good effort on both platforms. While some drops are disappointing, they seem rare enough to not be a big deal and is certainly more playable than the previous release.

Oh I got you. I wonder if reducing the OS's footprint might help.
The only likely solution is if they open up more cores to devs but that would require increased multithreading on devs' part.
 
This game exists on the PS3/360... not CPU monsters themselves, especially the PS3 (single PPU).

The bottlenecks probably resides elsewhere and come down to optimization, not something simple like CPU load. It's not like simple pathfinding for drivers and pedestrians need a lot of CPU.

Yeah, but they upped the amount of cars and pedestrians in these versions. I'm not a programmer, but I can't imagine this didn't have an effect on CPU usage.
 
Seriously with the sub 30fps stuff on current gen. Its really ridiculous. Shows you how underpowered both systems are, and particularly how crappy the CPUs are. And this is an upscaled remake to boot.

There is no way they make it past 5 years for this gen. The hardware can't support it.

No it just shows that ports that aren't designed for the strengths of the system won't perform optimally.
 

Marlenus

Member
It's really hilarious that everyone is going with this "CPU bound" thing now.

Yea it is just faulty analysis.

It is just as easy to say that MS have improved their API to such a degree that it now outperforms the PS4 API in certain tasks. Now I have no basis to believe that to be the case but it is just as valid an argument as this CPU bound nonsense.
 

bj00rn_

Banned
I'd really love to know who started this myth. You can't "lock" a frame rate. You can cap it so it doesn't go over 30, but it can always be less than that. Especially in a game that isn't heavily scripted.

But that's not really how we usually define "locked" framerate around here. "Locked framerate" usually means that the developer optimize, test and ensure the framerate is above 30 fps (f.ex.) at all times, and then vsync/cap it.
 

iceatcs

Junior Member
Not bad, even with upgrade graphics, the stress test wasn't bad. Maybe PS4 winning here with those better lighting/filters and more foliages but still ahead with the stress test.

DF has separated the result, to pick what you like.
 

dark10x

Digital Foundry pixel pusher
Weird, and honestly disappointing to see a game from last gen (albeit maybe the biggest game from last gen) still not hit a rock solid 30 fps. Really wish they would cut back on the effects to keep the frame-rate stable. At least give us the option to do so.
The performance video is designed to showcase the worst case scenarios, though. Having put hours into the game I can safely say that slowdown is barely an issue. The game holds a very steady frame-rate the vast majority of the time (on PS4 at least). It's doing some pretty impressive effects as well, though, which I wouldn't want to see cut as they add a tremendous amount to the visuals.

God of War remastered didn't run at a locked frame rate either. It had some deviation from 60.
Don't forget the original version of ZOE2. An absolute travesty. The fact that HexaDrive came along and reworked the whole thing into something infinitely superior should show how important optimization really is.
 
But that's not really how we usually define "locked" framerate around here. "Locked framerate" usually means that the developer optimize, test and ensure the framerate is above 30 fps (f.ex.) at all times, and then vsync/cap it.

and that's a lot easier to do in more scripted games or fighters, but in an open world game, I would imagine a lot of sacrifices would have to be made to ensure frame rate never goes below 30fps. It seems pretty solid in both versions.

Not bad, even with upgrade graphics, the stress test wasn't bad. Maybe PS4 winning here with those better lighting/filters and more foliages but still ahead with the stress test.

Also this they were actively trying to break the frame rate.
 

Rflagg

Member
I only watched the ps4 side, because I don't care about console war stuff, but I am not sure I will ever notice this while playing I wouldn't have noticed in the video if not for the numbers showing it. I beat this great game on ps3 and that dropped more frames and a lot more often so I am excited to play again with the huge visual upgrade and more stable frame rate. I think the remake/upgrade whatever the heck you want to call it compares well with any ps4/xbox one gen game out so far.
 

goonergaz

Member
The performance video is designed to showcase the worst case scenarios, though. Having put hours into the game I can safely say that slowdown is barely an issue. The game holds a very steady frame-rate the vast majority of the time (on PS4 at least). It's doing some pretty impressive effects as well, though, which I wouldn't want to see cut as they add a tremendous amount to the visuals.

it's all about the balance ;)
 
Top Bottom