• 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.

Substance Engine benchmark implies PS4 CPU is faster than Xbox One's

jonabbey

Member
I kind of doubt that Sony up-clocked their CPU, but having lower virtualization overhead makes sense, as does efficiencies with the memory system, the compiler, etc. The CPU may not be otherwise identical to the X1's, also, depending on the time line for each apu's development.

Maybe we'll find out some day.
 
I kind of doubt that Sony up-clocked their CPU, but having lower virtualization overhead makes sense, as does efficiencies with the memory system, the compiler, etc. The CPU may not be otherwise identical to the X1's, also, depending on the time line for each apu's development.

Maybe we'll find out some day.
The benchmark shows that at the very least, PS4's CPU is running at 1.75 GHz, just like XBone's, and it's likely to be running at 2 GHz.

Now, maybe it always ran at this speed and so this isn't technically an upclock, but it's certainly faster than we'd been led to believe.
 
Obviously the xbox was offline.

I'm sure when connected to the cloud we shall see different results with the Xbox CPU being infinitely faster than the PS4 CPU.

You would need at least a 112 Mbps internet connection to match the PS4 performance with the cloud. You would need close to gigabit fiber if the PS4 is using all its cores instead of just one.
 

Brohan

Member
I'm thinking the PS4 CPU runs at the same speed as the XB1 CPU except the PS4 has only 1 core reserved for OS while the XB1 has 2 cores reserved for OS.

Just a guess though.
 

Respawn

Banned
There was no evidence PS4 has a better CPU at all before this. Sony could have said something..
Surely the games proved this. Why would this new Sony say anything when certain sites and so called journalist already pledged allegiance? So many of you fall for these companies bullshit. The star this gen is Microsoft.
 

Respawn

Banned
All this talk about next gen consoles and the power of the hardware or lack thereof, yet when I look at games like Killzone and Ryse I tell myself...who gives a shit when the games look this good.
Good is one thing but what about performance? Ryse is a candidate for looks but lacking in performance.
 

KKRT00

Member
You certainly are shitting up this thread defending Ryse's framerate. Are you and Senj competing or something?

Sure, its my fault that thread derailed, because i want to fix misinformation in already derailed thread...

----------
It was Ryse.

Posts are very clear and simple, if You dont understand what i meant, thats Your problem.

Indicating that game runs in 60fps in the least complex situation and then comparing to the most heavy situation from other game was bullshit.
Ryse does not drop to 16 with few character on screen. KZ:SF does not run in 60fps in any scene with any complexity or view distance.
Holding 16 fps drop in Ryse as any indicator is stupid, because it was drop in the most heavy scenario in this game. Show me KZ:SF framerate analysis, where You fight 20 npcs with several explosions and we'll talk about 60fps or even 30fps. I could bet it drops to 20s in such a case [which is completely understandable].

If You still cant understand this, You'll land in my ignore list.
 

Vizzeh

Banned
I kind of doubt that Sony up-clocked their CPU, but having lower virtualization overhead makes sense, as does efficiencies with the memory system, the compiler, etc. The CPU may not be otherwise identical to the X1's, also, depending on the time line for each apu's development.

Maybe we'll find out some day.

Its not at upclock. The jaguars run at 2ghz on the 28nm chips... Only reason to "UNDERCLOCK" is for thermal reasons, if the PS4 is stable and cooling well, there is no reason to "underclock" to 1.6
 

KageMaru

Member
I'm thinking the PS4 CPU runs at the same speed as the XB1 CPU except the PS4 has only 1 core reserved for OS while the XB1 has 2 cores reserved for OS.

Just a guess though.

I don't think this is the case or GG would have used more than 6 cores for KZ.

It could be a number of things, such as performance reserved for the Kinect. Which makes this thread rather silly.
 

AgentP

Thinks mods influence posters politics. Promoted to QAnon Editor.
I don't think this is the case or GG would have used more than 6 cores for KZ.

It could be a number of things, such as performance reserved for the Kinect. Which makes this thread rather silly.

So if Kinect takes a huge amount of overhead, that is not relavent? To developers, the reason is not as important as the performance. But it is silly right?

Such a reversal from the days when they announced a clock bump and said how many games were CPU limited, but now it is all silly.
 

RoboPlato

I'd be in the dick
I don't think this is the case or GG would have used more than 6 cores for KZ.

It could be a number of things, such as performance reserved for the Kinect. Which makes this thread rather silly.

We don't know how many cores they used in the final game yet. The February reveal was what that info is from and things could have changed since then.
 

TTSupra

Banned
Sure, its my fault that thread derailed, because i want to fix misinformation in already derailed thread...

----------


Posts are very clear and simple, if You dont understand what i meant, thats Your problem.

Indicating that game runs in 60fps in the least complex situation and then comparing to the most heavy situation from other game was bullshit.
Ryse does not drop to 16 with few character on screen. KZ:SF does not run in 60fps in any scene with any complexity or view distance.
Holding 16 fps drop in Ryse as any indicator is stupid, because it was drop in the most heavy scenario in this game. Show me KZ:SF framerate analysis, where You fight 20 npcs with several explosions and we'll talk about 60fps or even 30fps. I could bet it drops to 20s in such a case [which is completely understandable].

If You still cant understand this, You'll land in my ignore list.

LOL your the one that quoted me with some KZ crap after I was comparing Ryse's supposly "100" units and its descent graphics to DR3's "100" units and cartoon graphics.

You should settle down also, you seem frantic to reply and extremely nervous with all your random word capitalization errors.
 
1. You are basing the comparison on launch games built on changing hardware that are often rushed out in time.

2. There are several sources of evidence that show that X1's tools were a mess before launch whereas the PS4 had more mature tools.

3. A CD Project Red dev came out and said that MS ends up being more open about their tools and shows devs more shortcuts later on in the generation whereas people have to create a network of devs in order to find the best ways to develop for Sony hardware.

4. The difference in the end, in my eyes, is probably going to be 900p vs. 1080p with the same effects. AC4 was like this, for one.

5. People usually don't count this but Kinect Sports Rivals is one of the nicest looking games on either console, and its running at 1080p on X1 whereas before they were planning to run at 720p, showing that the tools increased in quality enough to let them hit a better target.

On the bolded, not 100% the same effects. The PS4 version was patched to 1080p + received a much better AA solution. The XB1, on the other hand, was not able to receive such an uograde(stuck at 900p with some bad aliasing).
 
Xbox One games can improve graphically relative to PS4 to the extent hardware allows. Just the facts.

What does Xbox have left, btw? Maybe a better audio chip? One extra display plane scaler?
 
It could be a number of things, such as performance reserved for the Kinect. Which makes this thread rather silly.
Again, no. Kinect reserves 10% of the GPU. This thread is about the CPU, where the OS is rumored to reserve two cores on XBone, but only one core on PS4.

Regardless, this thread is about CPU power available to game devs. PS4 has more available; at least 17% more, and likely 33% more. Why it offers more power is interesting, though not technically relevant.
 
Posts are very clear and simple, if You dont understand what i meant, thats Your problem.

Indicating that game runs in 60fps in the least complex situation and then comparing to the most heavy situation from other game was bullshit.
Ryse does not drop to 16 with few character on screen. KZ:SF does not run in 60fps in any scene with any complexity or view distance.
Holding 16 fps drop in Ryse as any indicator is stupid, because it was drop in the most heavy scenario in this game. Show me KZ:SF framerate analysis, where You fight 20 npcs with several explosions and we'll talk about 60fps or even 30fps. I could bet it drops to 20s in such a case [which is completely understandable].
I'd bet Ryse would still run like shit with less characters on screen.

Oh and lets pretend Ryse doesn't drop frames with just two characters on screen.

If you still cant understand this, You'll land in my ignore list.
We'd all be better off if you put everyone on it.
 
I'd bet Ryse would still run like shit with less characters on screen.

Oh and lets pretend Ryse doesn't drop frames with just two characters on screen.

We'd all be better off if you put everyone on it.
Fucking A. We get that you have some weird stiffy for Crytek, KKRT00, but do you have to shit up threads like this? Honest question man, do you work for them? I never see you bad mouth anything they've done, always building up their work while putting down others'. Give it a rest.
 

flying dutchman

Neo Member
I didnt bring up Ryse to compare it to Killzone, I was making the point that it is a stunning looking game that was developed with Jaguar as its CPU. So obviously its more than capable and the best is yet to come.
 

Respawn

Banned
PS4 is like "You ain't got no wins in mi casa"


Just snatching every little gleam of hope out of the Xbox One soul.


Kinect & the audio chip is the only hardware advantage that the Xbox One have left & MS better do something to show them off.
Audio chip in PS4 is also superior I think. AMD's latest.
 

twobear

sputum-flecked apoplexy
1080P is 44% more pixels than 900P & PS4 has 56% more GPU power for games than Xbox One.


Xbox GPU power for games is 1.179 TFLOPS (1.31TFLOPS - 10% for Kinect/Snap)

PS4 GPU power for games is 1.843 TFLOPS ( has a secondary chip to help with the OS with it's own memory)


And this is without throwing in the GDDR5 advantage.

Ryse could have ran 1080P on the PS4.

I'm sure it could have done, but 60fps too? No.
 

viveks86

Member
I don't get it :(

tumblr_mkwlbj0bh91s7c8hwo1_500_zps6605e80a.gif
 

Shin

Banned
This thread is gonna be glorious despite the shitty specs of both machines (shitty to me because there's a lot of sacrifice in order to get that 1080p experience with everything maxed out).
Nevertheless it should make for some good reading, on that note it's kinda nice to find something new every other day about these consoles.
 

ElTorro

I wanted to dominate the living room. Then I took an ESRAM in the knee.
Holding 16 fps drop in Ryse as any indicator is stupid, because it was drop in the most heavy scenario in this game.

I played the game, and the occasions where you notice a framerate drop are indeed rare. It's not remotely comparable to games like, for instance, the last-gen console-versions of the Assassin's Creed games.

Ryse could have ran 1080P on the PS4.

Yes, that's likely. Crytek's comment on a hypothetical PS4 version was clearly a thinly veiled attempt to not stab Microsoft in the back.
 

Skeff

Member
Makes you wonder how these debates would go if Sony had gone with only 4GB of GDDR5.
At least then it wouldn't be a one-sided spec victory.

Would have been very interesting, by some accounts the 4GB GDDR5 would have been slightly faster than the 8GB GDDR5, so it would have had more bandwidth, but much less ram to use for games, it would have been taking it on a game by game basis regarding which multiplatform version was the best and the PS4 OS would have likely been less impressive.
 

ElTorro

I wanted to dominate the living room. Then I took an ESRAM in the knee.
Would have been very interesting, by some accounts the 4GB GDDR5 would have been slightly faster than the 8GB GDDR5, so it would have had more bandwidth, but much less ram to use for games, it would have been taking it on a game by game basis regarding which multiplatform version was the best and the PS4 OS would have likely been less impressive.

Developers can really be thankful to have relative parity regarding the amount of memory and CPU power, since those factors can influence the actual gameplay more than GPU power. The PS4 has the edge on both, but at leasts its in the same ballpark.
 

DC R1D3R

Banned
Developers can really be thankful to have relative parity regarding the amount of memory and CPU power, since those factors can influence the actual gameplay more than GPU power. The PS4 has the edge on both, but at leasts its in the same ballpark.

So everybody wins? Who'da thought it?
 

delta25

Banned
Sure, its my fault that thread derailed, because i want to fix misinformation in already derailed thread...

----------

.


You are aware that it was you and two others that really stirred the pot into shit, right? It's always about "misinformation", yet 16 pages in and still, no one gives a flying fuck what you have to say. Here's a good piece of advice for you...move on, you're really not doing this thread any justice with your "fixing".
 

Calabi

Member
It didnt make sense to me that the PS4's cpu was running at the default 1.6(who leaves things at the default?). They pushed all the other hardware so why wouldnt they the cpu as well?

I'm betting its 2.0ghz. It seems to run hotter than the Xbox one.
 

Arkam

Member
Both CPUs are ass. This is kind of a non-story, so I don't see much controversy.


Way Harsh... though pretty much true. But you cant stop people on here of taking .0001% difference and champion it as though it is the end all be all.
 

Sean*O

Member
Yeah Ryse could run 1080p on PS4, maybe shore up the framerate drops a bit at 1080p, but not a chance of 1080p 60 fps. It's about 1.5x more powerful, not 2-3x.

Hard to say. COD is running 2x better on PS4. Stuff I've seen of Ryse doesn't look that complicated, just really good texture work.
 

Skeff

Member
Way Harsh... though pretty much true. But you cant stop people on here of taking .0001% difference and champion it as though it is the end all be all.

If the information in the OP is correct then the PS4 is stronger per core, If the rumor about 1 CPU core being reserved on PS4 is also correct then that leaves the PS4 having about 33% more CPU available for games than XB1.

33% != .0001%
 

Gestault

Member
It makes me frown to see people thinking higher-than-expected CPU performance on the PS4 is anything but a good thing. It has no bearing on the XB1. The XB1 is just as fast as it ever was, so there's no need for frustration.
 
Yeah Ryse could run 1080p on PS4, maybe shore up the framerate drops a bit at 1080p, but not a chance of 1080p 60 fps. It's about 1.5x more powerful, not 2-3x.

resolution is likely mainly due to ROPs/Memory and res does not scale linearly with shader count

considering the PS4 has both greater effective bandwidth at larger pixbuff sizes and double the ROPs - I'm pretty sure they'd be able to hit 1080p 55 if not 60


It makes me frown to see people thinking higher-than-expected CPU performance on the PS4 is anything but a good thing. It has no bearing on the XB1. The XB1 is just as fast as it ever was, so there's no need for frustration.

see, the thing is GPU cost is easier to scale down than CPU cost without affecting gameplay as much

with the GPU you can turn off, fake or scale down small but heavy effects not very important to the gameplay like better explosions/particles/accurate reflections/ssao/sss etc.

with the CPU you can scale down audio, interactivity, physics, graphics and the like

all this combined will mean that cross platform multiplayer will be even more difficult as the balance may be off if we have an wide open map and the XB1 can only render at 900p so seeing people at distance may be harder or frame rate may be less stable so they miss more/react slower or that they don't see some reflection effect or directional sound when the enemy are behind a wall

they're all small things but they add up
 

imt558

Banned
Show me KZ:SF framerate analysis, where You fight 20 npcs with several explosions and we'll talk about 60fps or even 30fps. I


You talk bullshit. Even with 2 enemies on the screen, Ryse has framedrops. Yeah, trough entire game there is 20 enemies in every scene in Ryse. C' mon, dude. In Killzone Shadow Fall, Chapter 3 on ISA space cruiser Cassandra, after the reactor is on, you must get away from cruiser. In one scene there is 7 Helghast to fight against. Unlike NPC's in Ryse, those Helghast are not naked, those Helghast wearing armor suit which is veeeery detailed with several materials , they don't have some stupid sword, they have guns ( some guns with flaslights to ) which creates lightning after each fire, Helghast throwing grenades on you. I return fire. Oh, i forgot. I have one NPC to - the OWL, OWL return fire to, OWL make blast which stuned enemies.... Of course, there is framedrops to in that scene, but there is so much going on on screen, much more than some scene in Ryse. And some people said that Ryse is graphically better than KZ. Sure, sure.... In another world.
 

3Dawg

Banned
You talk bullshit. Even with 2 enemies on the screen, Ryse has framedrops. Yeah, trough entire game there is 20 enemies in every scene in Ryse. C' mon, dude. In Killzone Shadow Fall, Chapter 3 on ISA space cruiser Cassandra, after the reactor is on, you must get away from cruiser. In one scene there is 7 Helghast to fight against. Unlike NPC's in Ryse, those Helghast are not naked, those Helghast wearing armor suit which is veeeery detailed with several materials , they don't have some stupid sword, they have guns ( some guns with flaslights to ) which creates lightning after each fire, Helghast throwing grenades on you. I return fire. Oh, i forgot. I have one NPC to - the OWL, OWL return fire to, OWL make blast which stuned enemies.... Of course, there is framedrops to in that scene, but there is so much going on on screen, much more than some scene in Ryse. And some people said that Ryse is graphically better than KZ. Sure, sure.... In another world.

Joke post?

It's a joke post right?
 
Top Bottom