• 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.

DF: Xbone Specs/Tech Analysis: GPU 33% less powerful than PS4

I'll wait for the games to do that talking and not hardware spec sheets. Sure it might be indicative for now but we haven't seen what either console can really do yet
 

andycapps

Member
The RSX graphics chip in the PS3 had around 80% more raw power than the Xbox 360's...look how that turned out. The Xbox (Original) had ALOT more power than PS2...look how that turned out. Sony will push the PS4...but nobody else will.

To people saying Xbox One = WiiU....NO. This is a modern architecture with raw performance at 1.3TFLOPS + cloud off-load Vs WiiU's archaic architecture running 300-400 TFLOPS. They're not even close.

Also...apparently nobody is taking into account CPU.

Also, almost nothing is known about real world performance - ie. Direct X 11 (exclusive to the Xbox One) will allow many shortcuts to developers...it's apparently very powerful.

incredulous.gif
 
I just read elsewhere that some people are saying "there's strong evidence" the PS4's GPU uses SISD instructions while One's GPU uses SIMD.

Is something like that even possible?
 
I just read elsewhere that some people are saying "there's strong evidence" the PS4's GPU uses SISD instructions while One's GPU uses SIMD.

Is something like that even possible?


I guess they're talking about VLIW4 vs GCN, but I believe the X1 is confirmed (or very nearly confirmed) to be using GCN, just like the PS4.
 

Zukuu

Banned
Will 33% even be noticeably different? Feels like 33% isn't a big difference..
Instead of 3 PS3 combined you suddenly only have 2. It does make a difference. The question is if the devs will make a difference. Multiplats could run in 60FPS on PS4 and only 30FPS on XBone for instance.
 

kitch9

Banned
The RSX graphics chip in the PS3 had around 80% more raw power than the Xbox 360's...look how that turned out. The Xbox (Original) had ALOT more power than PS2...look how that turned out. Sony will push the PS4...but nobody else will.

To people saying Xbox One = WiiU....NO. This is a modern architecture with raw performance at 1.3TFLOPS + cloud off-load Vs WiiU's archaic architecture running 300-400 TFLOPS. They're not even close.

Also...apparently nobody is taking into account CPU.

Also, almost nothing is known about real world performance - ie. Direct X 11 (exclusive to the Xbox One) will allow many shortcuts to developers...it's apparently very powerful.

Pro-tip.

If you don't want to look really stupid on the internet, don't post stuff like this.

Fucking jesus wept.
 

onQ123

Member
Will 33% even be noticeably different? Feels like 33% isn't a big difference..

That's 33% of PS4's GPU power but 50% of Xbox One's GPU.

you only use 33% when you are trying to gauge Xbox One's power from looking at the PS4 but if you're trying to gauge the PS4 power by looking at the Xbox it's going to be 50% more.
 

Mitchings

Banned
Has anyone heard anything yet about the Compute abilities of the Xbox One's GPU relative to..

PS4's 64-Compute Queues (8-ACE's x 8-CL's)

&

[Current-Gen AMD GPU] PC's 4-Compute Queues (2 ACE's x 2CL's)

..?
 
The RSX graphics chip in the PS3 had around 80% more raw power than the Xbox 360's...look how that turned out. The Xbox (Original) had ALOT more power than PS2...look how that turned out. Sony will push the PS4...but nobody else will.

To people saying Xbox One = WiiU....NO. This is a modern architecture with raw performance at 1.3TFLOPS + cloud off-load Vs WiiU's archaic architecture running 300-400 TFLOPS. They're not even close.

Also...apparently nobody is taking into account CPU.

Also, almost nothing is known about real world performance - ie. Direct X 11 (exclusive to the Xbox One) will allow many shortcuts to developers...it's apparently very powerful.

nwuVK9x.gif
 

CLEEK

Member
Will 33% even be noticeably different? Feels like 33% isn't a big difference..

Remember, due to the way percentages work, the Xbone is 33% less powerful than the PS4, and the PS4 is 50% more powerful than the Xbone.

And these figure were before it was know that the Xbone reservces 10% of its GPU resources, so the actual figuires show the PS4 is 60% more powerful (~1.8FT vs ~1.1TF). This is a substantial amount on paper. But in real world gaming?

Back before the reveals when we were all discussing the rumoured specs, there were links (that's I'm trying to find again) that showed real gaming performance for AMD cards with similar specs as the XBone and PS4.

18CUs and GGDR5 vs 12CU and GDDR3.

The results varied game by game, but went from games running 50% higher frame rate on the better card, to some instances where the frame rate was double.

So it is technically possible that you could have the situation where Xbone games have the same resolution and effects as the PS4 game, but run at half the frame rate to accommodate it. Or to keep the frame rate, it will have to greatly scale back resolution/textures/effects.
 
Also, almost nothing is known about real world performance - ie. Direct X 11 (exclusive to the Xbox One) will allow many shortcuts to developers...it's apparently very powerful.


Don't you need beefy hardware to get the most out of DX11? And isn't that the one thing Xbone doesn't have?
 

CLEEK

Member
Here we go:

http://www.anandtech.com/bench/Product/536?vs=549

This shows the performance difference between the two card most similar in FLOPS to the PS4 and Xbone's APU. The Xbone equivalent will perform better than the actual console, due to the 7770 having GDDR5 rather than the much slower GDDR3 used in the console.

So 33%/50% (however you say it) difference does have noticeable differences even in a PC. In a console, where games are coded to the metal to max out the performance benefits, the differences will be greater.

For the games that do push the PS4, the difference over the Xbone should be pretty major. GDDR5 aside, it has an extra ~700GFlops over the Xbone (that's like two WiiUs duct taped together, or two thirds of an Xbone).
 

TAJ

Darkness cannot drive out darkness; only light can do that. Hate cannot drive out hate; only love can do that.
JoeTheBlow's tag... damn.
Herman Munster was right. Sometimes dead is better.
 

silaslang

Banned
In other words, Sony got lucky GDDR5 production went smoothly (or they just took a big risk).

Will be interesting to see if Durango GPU will have boosted it's clockspeeds, as it's not confirmed yet. If it is, then performance gap with PS4 will not be as big. Although clock speeds can only do so much and it's very bad for power performance.

I personally doubt it will be anything else than the 1.2 TFLOPS as has been rumoured.

edit:


wat the actual fuck
The hd is not replaceable but an external hd has all the functions available. As in you can install and run games on it. Plug it in an extend storage.
 

onQ123

Member
Here we go:

http://www.anandtech.com/bench/Product/536?vs=549

This shows the performance difference between the two card most similar in FLOPS to the PS4 and Xbone's APU. The Xbone equivalent will perform better than the actual console, due to the 7770 having GDDR5 rather than the much slower GDDR3 used in the console.

So 33%/50% (however you say it) difference does have noticeable differences even in a PC. In a console, where games are coded to the metal to max out the performance benefits, the differences will be greater.

For the games that do push the PS4, the difference over the Xbone should be pretty major.

Also the HD 7770 is clocked at 1GHz & pushing out 1.28TFLOPS a little over the Xbox One 1.23 TFLOPS while the HD 7850 is pushing out 1.76 TFLOPS a little less than the PS4's 1.84TFLOPS & the memory bandwidth is 153.6GB/s a little slower than the PS4 memory bandwidth.
 

Painraze

Unconfirmed Member
2013 will go down in history as the year fanboys have truly lost their minds. I never thought it would get worse than the PS3/360 battle years ago... I was way off.
 

i-Lo

Member
Also the HD 7770 is clocked at 1GHz & pushing out 1.28TFLOPS a little over the Xbox One 1.23 TFLOPS while the HD 7850 is pushing out 1.76 TFLOPS a little less than the PS4's 1.84TFLOPS & the memory bandwidth is 153.6GB/s a little slower than the PS4 memory bandwidth.

One thing to remember is that the memory bandwidth for PS4 be actually around ~156GB/sec given the CPU generally utilizes <20GB/s.

Guess it's time for me to leave this behind again...

Extra Credits: Why Console Specs Don't Matter

You come into technical discussion thread and this is what you leave behind. Kind of seems like thread whining.

More to the point, given the near identical base design followed by closely matching price point and release date, it does matter.
 

CLEEK

Member
Also the HD 7770 is clocked at 1GHz & pushing out 1.28TFLOPS a little over the Xbox One 1.23 TFLOPS while the HD 7850 is pushing out 1.76 TFLOPS a little less than the PS4's 1.84TFLOPS & the memory bandwidth is 153.6GB/s a little slower than the PS4 memory bandwidth.

Yeah, it's not perfect (doesn't factor in the extra CPU core available for PS4 games, the RAM size and bandwidth difference, the eDRAM etc), but it's a pretty good indication using actual metrics rather than guesses.

Point is, the PS4 has a non-trivial advantage over the Xbone. And unlike every prior gen, due to the PS4 and Xbone being based on the exact same architecture, it's not like the Xbone makes up for the differences in other areas. The PS4 is just the better box all round.
 

onQ123

Member
Yeah, it's not perfect (doesn't factor in the more CPU cores available for PS4 games, the RAM size and bandwidth difference, the eDRAM etc), but it's a pretty good indication using actual metrics rather than guesses.

Point is, the PS4 has a non-trivial advantage over the Xbone. And unlike every prior gen, due to the PS4 and Xbone being based on the exact same architecture, it's not like the Xbone makes up for the differences in other areas. The PS4 is just the better box all round.

Yeah been telling people for awhile that this is way bigger than the PS3 & Xbox 360 difference.
 

nib95

Banned
Here we go:

http://www.anandtech.com/bench/Product/536?vs=549

This shows the performance difference between the two card most similar in FLOPS to the PS4 and Xbone's APU. The Xbone equivalent will perform better than the actual console, due to the 7770 having GDDR5 rather than the much slower GDDR3 used in the console.

So 33%/50% (however you say it) difference does have noticeable differences even in a PC. In a console, where games are coded to the metal to max out the performance benefits, the differences will be greater.

For the games that do push the PS4, the difference over the Xbone should be pretty major. GDDR5 aside, it has an extra ~700GFlops over the Xbone (that's like two WiiUs duct taped together, or two thirds of an Xbone).

PS4's GPU is actually closer to a 7870 minus the clock speed (which is rumoured to be going up anyway). At the very least, somewhere in between the 7850 and 7870.

Here's how the 7770 stacks up to the 7870. As you mentioned, the reality is the XO is using slower DDR3 ram compared to this GPU, and a slower core clock, but I believe has more ROP's.

7770 (est Xbox One) vs 7870 (est PS4)

On average at 1080p, you're looking at almost double the frame rate for the PS4 equivalent, which makes me think a 720p vs 1080p, or 1080p 30fps vs 1080p 60fps difference, may actually be possible. At the very least there should be some evident graphical differences in multiplatform games unless shenanigans are afoot. Remember, architecturally they are extremely similar.
 

JaggedSac

Member
You guys are also leaving out the fact that 10% of the gpu is reserved for the app OS. At least I think I recall that.

Edit: Speaking about the XBone.
 

i-Lo

Member
PS4's GPU is actually closer to a 7870 minus the clock speed (which is rumoured to be going up anyway). At the very least, somewhere in between the 7850 and 7870.

Here's how the 7770 stacks up to the 7870. As you mentioned, the reality is the XO is using slower DDR3 ram compared to this GPU, and a slower core clock, but I believe has more ROP's.

7770 (est Xbox One) vs 7870 (est PS4)

I'd say a direct comparison is rather unfair by using their closest retail counterpart because the customizations should allow for much more (in terms of compute functionality and flexibility).
 

nib95

Banned
The RSX graphics chip in the PS3 had around 80% more raw power than the Xbox 360's...look how that turned out.

No it did not. Nvidia fluffed the numbers and pretended it did, but the reality was that Xenos (360's GPU) had the performance advantage over RSX (PS3's GPU).
 

nib95

Banned
Just to correct you on one thing SPE, you keep saying that the Xbox One uses GDDR3, but it actually uses DDR3 (a slower system usage variant).
 

Spongebob

Banned
PS4's GPU is actually closer to a 7870 minus the clock speed (which is rumoured to be going up anyway). At the very least, somewhere in between the 7850 and 7870.

Here's how the 7770 stacks up to the 7870. As you mentioned, the reality is the XO is using slower DDR3 ram compared to this GPU, and a slower core clock, but I believe has more ROP's.

7770 (est Xbox One) vs 7870 (est PS4)

On average at 1080p, you're looking at almost double the frame rate for the PS4 equivalent, which makes me think a 720p vs 1080p, or 1080p 30fps vs 1080p 60fps difference, may actually be possible. At the very least there should be some evident graphical differences in multiplatform games unless shenanigans are afoot. Remember, architecturally they are extremely similar.
7770 vs 7850 is a much better comparison.
 

Shayan

Banned
Guess it's time for me to leave this behind again...

Extra Credits: Why Console Specs Don't Matter

you will be seeing much bigger difference than what you with ps3 vs xbox and i expect most multiplat games to look/run significantly better too

1.8TF vs 1.1TF , 18CU vs 12 CU , 5 gig of slow DDR3 vs 7gig of ultrafast GDDR5. Both the machines are built with PC based architecture, so expect the difference to show from right at launch

Oh, so just ignore the ESRAM on Xbox One? How silly. I wonder why I ever bother to read any of this stuff lol.

that 32mb esram is slower and less efficient than GDDR5. DAT 170g/s bandwith is accessible to only 32mb ram (68g/s for ddr3 and 102g/s for esram itself). Nothing will magically bump the bandwidth of slow ddr3 from 68g/s
 

i-Lo

Member
Oh, so just ignore the ESRAM on Xbox One? How silly. I wonder why I ever bother to read any of this stuff lol.

Perhaps because you simply don't just add the two together trying to pass it off as a sustainable rate. That said, they should have mentioned its existence separately at the least.
 
Also consider the rumor that you can code to the metal on PS4 while you have to go through Microsoft's layers on Xbone.

The difference should easily be 50-100% better framerates for multiplats, or higher resolution/effects.

First party games developed ground up for PS4 should also show a pretty huge advantage, larger than what we saw last gen for sure.
 
Top Bottom