EDGE: "Power struggle: the real differences between PS4 and Xbox One performance"

We're enthusiasts on this forum so we compare enthusiast hardware. Not grandma's internet browser. Obviously. However being a modular setup, you can spend whatever you'd like and cater your gaming to your needs obviously. There's no need to go into a "what do I spend" debate because there are other threads for that. This one's comparing multiplatform performance. On enthusiast hardware, there really isn't a contest.



This will be my final post of this nature, because apparently we have segregated threads now (wtf IN&OUT? no need to go there - we're all friendly here on GAF) that don't warrant discussion on which multiplats are better where (which was the reason GAF user "Horp" brought PC users into this discussion). However a lot of your notions (such as manual tweaking - entirely optional) and interfaces are out of date. It's all just as easy now to operate your games and even hardware and drivers with (free) services like Steam. Modularity is a net positive, however, and you are correct that it's different. My only point was that it seemed a bit disingenuous to compare the same multiplats being compared here and made a huge fuss over in a 40 page thread about a (small in comparison) 500gf GPU difference when it's fairly easy to have a double or triple difference in performance should you choose to want to have the best multiplatform options.

There will be plenty of people purchasing GTA 5 shortly on systems with massively compromised performance/IQ in comparison to the PC build that Rockstar is showing in the commercials. And that's cool! It's totally fine. I've played plenty of subpar ports this generation, on and off the PC. But I don't pretend that I'm getting the best version when I play them, and I don't fool myself into thinking i'm playing the best version over small difference from one platform to another when I myself have chosen to play it on a console where compromise is in its very nature. When I play on the consoles, I accept things like tearing and subpar framerates and awful IQ, and I don't even see it after I get used to it. It's always going to be in second (or third, or 4th or whatever) place, and I don't proclaim that Call of Duty looks SO MUCH BETTER on the 360 than on the PS3 because of a slightly less subHD render. If I cared SO MUCH about how much better it looks from one minor difference to another, why would I limit myself to a choice between 2 subpar versions in comparison to the one that runs in 1080p/60 (or whichever resolution I happen to choose on my own)? I hope that sort of helps you understand my disconnect here.

Should you choose to limit yourself to a gaming box built by either Sony or MS instead of you, yes the PS4 will absolutely perform better. It will have better multiplatform games, and is a better value as a gaming box limited to that comparison. That much is not in question and nobody should question it at this point. But let's not pretend there aren't other gaming boxes either, running the same games. A box is a box is a box. It's just a means to an end. When the HDMI cable leaves the box, it does the same thing whether it's a $4000 BF4 PC or an Ouya.

Good post.
 
Despite the power difference, it'll only become an issue if PS4 sells a significant amount more than X1. Only then will we start seeing a performance gap as devs start using PS4 as the lead platform.

No, Sony's first party games will make the power difference apparent. Some third party devs might have games looking the same on both consoles, but as the generation goes on differences will show, just like this gen.
 
Despite the power difference, it'll only become an issue if PS4 sells a significant amount more than X1. Only then will we start seeing a performance gap as devs start using PS4 as the lead platform.

Not this lead platform shit again, I am sure both PS4 and XBone will be utilised as much as possible. If that was the case with lead platforms then all those PC ports shouldn't look much better than consoles, this has happened a few times where GTAIV was horribly optimised on PC, but on this occasion all three platforms are similar. I expect we will see a difference straight away at launch and continue to see a difference as the generation goes by.
 
So...from all this spec talk on GAF, is it safe to say that PS4's first party teams will develop games that show the differences in strength level to be comparable to say Xbox-PS2?

Because anything else, and I don't think people will notice that much. People did notice titles like God of War or Uncharted looking really amazing, but 360 still stomped PS3 in the U.S...

The difference this time is fairly bigger though.
 
Despite the power difference, it'll only become an issue if PS4 sells a significant amount more than X1. Only then will we start seeing a performance gap as devs start using PS4 as the lead platform.

The devs of Witcher 3 have already stated they have individual teams working on each platform version (XB1, PS4, PC) so no lead platform there

I think we'll see the difference in some games at launch but it will likely be small

Start to get bigger as time progresses though
 
I'm totally calm- there was zero yelling in my previous post. ;)

Also- I'm not shouting from the rooftops proclaiming that the difference between PS4 and XB1 is mathematically equal to that of PS2 and XBOX. I was replying to a post that had already made two specific comparisons, (NES vs. PS3 and Xbox vs. PS2) and kept them the same in my response for clarity's sake. EDIT: I think I chose my words poorly, though. My point was that if you noticed the difference between Xbox and PS2, you're going to notice the difference between XB1 and PS4, not that the difference was the same.

As for my reply to you- I wasn't commenting on the CG vs. Xbox part of the comment, I was commenting on the idea of using launch titles to gauge the potential of the new consoles versus looking at the specs.

Your point doesn't make sense though. The differences between the Xbox and PS2 were extremely noticeable due to an extreme gap between the two (a ~ 200% to 300% gap). You saying that someone who noticed such a gap will notice the one between the XB1 isn't entirely true. A more apt comparison would be to say that someone who notices the difference between a GC and an Xbox ( or maybe a Wii and a GC since that's what the difference actually is mathematically if a poster on here is correct) will notice the difference between and XB1 and a PS4.

As for specs bit, I think if you can get games like Ryse/Forza as launch titles that look as good as KZ/Infamous from a console that is supposedly tricky to develop for and needs tinkering/time to figure out, compared to a console that is much easier and much more straightforward to develop for, then I don't think the gap will be that big. That's essentially what I was saying. The XB1 is still a very capable machine despite being the weaker of the two. People acting like we are going to be getting a PS2/ogXbox gap between the two need to re-evaluate their critical thinking skills.
 
So...from all this spec talk on GAF, is it safe to say that PS4's first party teams will develop games that show the differences in strength level to be comparable to say Xbox-PS2?

Because anything else, and I don't think people will notice that much. People did notice titles like God of War or Uncharted looking really amazing, but 360 still stomped PS3 in the U.S...

Considering what Sony Santa Monica and Naughty Dog did with 6-year old (almost 7) hardware, I think so.
 
Your point doesn't make sense though. The differences between the Xbox and PS2 were extremely noticeable due to an extreme gap between the two (a ~ 200% to 300% gap). You saying that someone who noticed such a gap will notice the one between the XB1 isn't entirely true. A more apt comparison would be to say that someone who notices the difference between a GC and an Xbox ( or maybe a Wii and a GC since that's what the difference actually is mathematically if a poster on here is correct) will notice the difference between and XB1 and a PS4.

Just going off percentages isn't really a good way to look at this. IIRC, the power gap is wide enough that it's literally the total power of Xbox One + 360 + ps3 equals the power of the ps4.
 
Not this lead platform shit again, I am sure both PS4 and XBone will be utilised as much as possible. If that was the case with lead platforms then all those PC ports shouldn't look much better than consoles, this has happened a few times where GTAIV was horribly optimised on PC, but on this occasion all three platforms are similar. I expect we will see a difference straight away at launch and continue to see a difference as the generation goes by.

Well my comment had in mind Edge's quote about multiplatform performance becoming more of a political issue; forcing a parity between each so as to not ruffle feathers. The only way around this is a large enough install base gap. Then devs will stop caring about the political side of it.
 
Well my comment had in mind Edge's quote about multiplatform performance becoming more of a political issue; forcing a parity between each so as to not ruffle feathers. The only way around this is a large enough install base gap.

That would be new for the industry. Previously, publishers didn't seem to care, and one console regularly out preformed the other.
 
Well my comment had in mind Edge's quote about multiplatform performance becoming more of a political issue; forcing a parity between each so as to not ruffle feathers. The only way around this is a large enough install base gap.

I personally think this is bullshit, every generation in history from Amiga , and NES, SNES and Genesis to PS2 and Xbox to PS3 and 360, devs didn't give a fuck about having versions comparable to each other, they simply did what they can with the systems they were working on, even if necessary version look superior to the other.
 
There's a bigger chance that MS pressures Digital Foundry than actual developers with respect to parity between multiplat titles.

DF afaik is the only major, credible site that compares titles between platforms in such detail so they will hold a lot of cards next gen. I mean who else is there to challenge their claims if they tell us that, in fact, there is only a few frames difference in the opening scene of BF4? Who else benchmarks console titles so accurately? Lens of Truth?!
 
Despite the power difference, it'll only become an issue if PS4 sells a significant amount more than X1. Only then will we start seeing a performance gap as devs start using PS4 as the lead platform.

Why are you assuming the X1 will be the lead platform for developers?
 
it's true as far as flops. Just add up the respective systems. gflops are not the end-all-be-all metric for game consoles though

So basically what he is saying is that because the FLOPS difference between the two is a 360+PS3, somehow the power gap between the two is an XB1+360+PS3? Is that even a logical conclusion? Doesn't seem like it to me.
 
All of a sudden PC thread?

I didn't see the guy mention his FalconNW machine in the article... must have been skimming.

Moving on...
Infamous SS already smokes everything I've seen come out of MS camp so far and it's a launch window open world game. Sony's 1st party consistently throttles every competitor in just about every aspect. I can not imagine how ridiculous some of their studio work is going to look in a years time.
 
Your point doesn't make sense though.

Only because you're hung up on the numbers and you're trying to apply everything literally instead of putting that post in the context of the post it was replying to.

As for specs bit, I think if you can get games like Ryse/Forza as launch titles that look as good as KZ/Infamous from a console that is supposedly tricky to develop for and needs tinkering/time to figure out, compared to a console that is much easier and much more straightforward to develop for, then I don't think the gap will be that big.

Launch games. How do they work?

I'll make an analogy.

DISCLAIMER: Now, bear in mind, this is just a rough example. I'm not literally saying that the XB1 and the PS4 are the exact equivalent of a Chevy Cavalier and a Ford Mustang. Please don't post replies explaining to me that the percentages don't add up. I'm well aware of that. Again- this is just to illustrate a point.

A Chevy Cavalier and a Ford Mustang are going to participate in a race. Prior to lining up at the start line and going full blast, though- the two cars are going to do a single lap around the track to introduce themselves to the crowd, and to give the drivers a chance to get used to how their cars handle. Both cars will go maybe 70 or 80 mph for this lap.

That's what a console launch is in this analogy. It's the warm up lap. Don't assume that just because the Cavalier went as fast as the Mustang in this lap that it's going to be able to keep up as the real race progresses.
 
It's not all about resolution and frame rates. With 50% more gpgpu you can have more physics, ai, and whatever else and the games could still be equals graphically.

MS is in a bad spot.
 
I'm surprised they haven't started banning people bringing PC!wars in here lol
Still. Can't wait for the release comparisons. Shit will be Interesting....
 
So basically what he is saying is that because the FLOPS difference between the two is a 360+PS3, somehow the power gap between the two is an XB1+360+PS3? Is that even a logical conclusion? Doesn't seem like it to me.

lol no what sangreal was saying and you agreed with is the following

FLOP-wise

PS4 = XB1 + X360 + PS3

That's it

Therefore

PS4 - XB1 = X360 + PS3 > 0

*Cloud power not included as it's total bullcrap
 
If people were fine buying ps3 versions over 360 this gen then the difference isn't gonna matter. Just be happy you're getting the games either way.

edit:
That's why I'm going ps4 and wii u only. Both have a strong first party, Microsoft seriously fucked up all there studios and have nothing really.

That's a much better reason.
 
Proof is in the pudding. PS3 was supposed to make 360 look like Xbox 1.5.

We all know how that turned out

Actually that statement was a reaction to how 360 games looked compared to Xbox games. PS3 games hadn't even been shown when the Xbox 1.5 comments were first being made. And this gen has a much more obvious difference in specs and dev comments.
 
Only because you're hung up on the numbers and you're trying to apply everything literally instead of putting that post in the context of the post it was replying to.



Launch games. How do they work?

I'll make an analogy.

DISCLAIMER: Now, bear in mind, this is just a rough example. I'm not literally saying that the XB1 and the PS4 are the exact equivalent of a Chevy Cavalier and a Ford Mustang. Please don't post replies explaining to me that the percentages don't add up. I'm well aware of that. Again- this is just to illustrate a point.

A Chevy Cavalier and a Ford Mustang are going to participate in a race. Prior to lining up at the start line and going full blast, though- the two cars are going to do a single lap around the track to introduce themselves to the crowd, and to give the drivers a chance to get used to how their cars handle. Both cars will go maybe 70 or 80 mph for this lap.

That's what a console launch is in this analogy. It's the warm up lap. Don't assume that just because the Cavalier went as fast as the Mustang in this lap that it's going to be able to keep up as the real race progresses.

Don't condescend me. I fully understand the point you are trying to make, and I don't necessarily disagree with it (with regards to launch games).

Your analogy is broken and selective because your analogy assumes that both cars are trying to maintain parity for an introductory show, then begin the race. This isn't true for first party launch games. Guerilla isn't trying to make a game that is on par with Ryse or Forza. Guerilla is trying to make an awesome game with the time and knowledge they have. So is Crytek, Remedy, and Turn10. So in essence, both cars are already racing, so to speak.

So to use a similar analogy, two cars are racing a total of 10 laps. One car is alleged to be 50% weaker in raw performance and is more exotic/difficult to drive. However, at the start of the first lap, both cars are almost neck and neck. You are saying that there is going to be a "massive" gap. I'm saying that the current gap on the track does not appear to be massive, and that the gap in the future will not be as massive as you are suggesting.

lol no what sangreal was saying and you agreed with is the following

FLOP-wise

PS4 = XB1 + X360 + PS3

That's it

Therefore

PS4 - XB1 = X360 + PS3 > 0

*Cloud power not included as it's total bullcrap

Right, I know that. What I am asking is whether a FLOPs ( as in instructions per second) gap of 360+PS3 (ie PS4=XB1+360+PS3) translates to a POWER gap of 360+PS3? I don't know a ton about tech stuff, but it seems misleading to make broad performance conclusions based on a single variable.
 
Can anyone explain to me how the Xbox One architecture is balanced? It looks like damn mess.

so many compromises have been made because of slow ass DDR3. Wasted silicon space on the APU, extra chips and memory pools all make for one messy unbalanced system in my opinion.

Sure, if the VGLeaks article on the PS4 architecture is accurate, the GPU in there is a 14 CU GPU with 4 extra CUs tacked on.

There's very little information available about those 4 extra CUs. Cerny was asked about this by DF and responded that they added a little more ALU than you normally would to encourage studios to use GPGPU for things like audio raycasting, etc.

If the 14+4 GPU is correct it may not have extra texture cache, L1 cache, L2 cache or TLB cache for those CUs. The PS4 works around that by supplying the Onion+ bus which gives the GPU a channel to memory at 20 GBps which doesn't use the L1 or L2 cache. Bypassing the caches means that the GPGPU performance would be slower on CUs using Onion+ as they have to wait for the writes to complete before caring on. Memory caches are used to buffer memory into fast on-chip RAM.

The Xbox One doesn't have these extra CUs but it does have some fixed function co-processors to do jobs like audio raycasting and moving memory, and it has a fast on-chip scratch pad which will have much less latency than writing to external memory (if the PS4's ROPs suffer from ~2x latency of the Xbox One, their ROP throughput will be similar when writing to ESRAM). The leaked PS4 GPU clock is 800mhz and the Xbox One is 853mhz so you've got 12@853 vs 14+4@800. I doubt anyone in this thread knows exactly how that will play out in the long term.

This is where the balanced argument comes from.
 
Don't condescend me. I fully understand the point you are trying to make, and I don't necessarily disagree with it (with regards to launch games).

Your analogy is broken and selective because your analogy assumes that both cars are trying to maintain parity for an introductory show, then begin the race. This isn't true for first party launch games. Guerilla isn't trying to make a game that is on par with Ryse or Forza. Guerilla is trying to make an awesome game with the time and knowledge they have. So is Crytek, Remedy, and Turn10. So in essence, both cars are already racing, so to speak.

So to use a similar analogy, two cars are racing a total of 10 laps. One car is alleged to be 50% weaker in raw performance and is more exotic/difficult to drive. However, at the start of the first lap, both cars are almost neck and neck. You are saying that there is going to be a "massive" gap. I'm saying that the current gap on the track does not appear to be massive, and that the gap in the future will not be as massive as you are suggesting.

Your analogy only works if both cars are currently in the process of being built and tuned.
 
Sure, if the VGLeaks article on the PS4 architecture is accurate, the GPU in there is a 14 CU GPU with 4 extra CUs tacked on.

It's just 18CUs, there are no hardware differences. The "14+4" is just an example of how you could schedule graphics and compute, if you want to.

Digital Foundry: Going back to GPU compute for a moment, I wouldn't call it a rumour - it was more than that. There was a recommendation - a suggestion? - for 14 cores [GPU compute units] allocated to visuals and four to GPU compute...

Mark Cerny: That comes from a leak and is not any form of formal evangelisation. The point is the hardware is intentionally not 100 per cent round. It has a little bit more ALU in it than it would if you were thinking strictly about graphics. As a result of that you have an opportunity, you could say an incentivisation, to use that ALU for GPGPU.

http://www.eurogamer.net/articles/digitalfoundry-face-to-face-with-mark-cerny

The Xbox One doesn't have these extra CUs but it does have some fixed function co-processors to do jobs like audio raycasting [...]

You can't do raycasting on the XBO's audio hardware. You access it via the XAudio2 API [1] which is for processing audio streams, not for casting rays on the geometry of the game world.

[1] http://msdn.microsoft.com/en-us/library/windows/desktop/hh405049(v=vs.85).aspx
 
Your analogy only works if both cars are currently in the process of being built and tuned.

My analogy would actually not make sense if both cars are still being built (since you can't race [not test, but race] an incomplete car?) What does this even mean? Please clarify that.

And we can pretty much say that we know what the first batch of games will look like, barring a few optimizations here and there. So in essence, the race has started. The consoles are roughly finalized and are being mass produced.
 
Don't condescend me. I fully understand the point you are trying to make, and I don't necessarily disagree with it (with regards to launch games).

Your analogy is broken and selective because your analogy assumes that both cars are trying to maintain parity for an introductory show, then begin the race. This isn't true for first party launch games. Guerilla isn't trying to make a game that is on par with Ryse or Forza. Guerilla is trying to make an awesome game with the time and knowledge they have. So is Crytek, Remedy, and Turn10. So in essence, both cars are already racing, so to speak.

So to use a similar analogy, two cars are racing a total of 10 laps. One car is alleged to be 50% weaker in raw performance and is more exotic/difficult to drive. However, at the start of the first lap, both cars are almost neck and neck. You are saying that there is going to be a "massive" gap. I'm saying that the current gap on the track does not appear to be massive, and that the gap in the future will not be as massive as you are suggesting.

You can't say how big that gap will be because we don't know. But once the drivers get accustomed to the track and their cars, there's a good chance we're gonna see that gap grow.

The devs quoted in this article have called the difference "significant" and "obvious." You don't say that unless there's a fairly decent difference.
 
Sure, if the VGLeaks article on the PS4 architecture is accurate, the GPU in there is a 14 CU GPU with 4 extra CUs tacked on.

There's very little information available about those 4 extra CUs. Cerny was asked about this by DF and responded that they added a little more ALU than you normally would to encourage studios to use GPGPU for things like audio raycasting, etc.

If the 14+4 GPU is correct it may not have extra texture cache, L1 cache, L2 cache or TLB cache for those CUs. The PS4 works around that by supplying the Onion+ bus which gives the GPU a channel to memory at 20 GBps which doesn't use the L1 or L2 cache. Bypassing the caches means that the GPGPU performance would be slower on CUs using Onion+ as they have to wait for the writes to complete before caring on. Memory caches are used to buffer memory into fast on-chip RAM.

The Xbox One doesn't have these extra CUs but it does have some fixed function co-processors to do jobs like audio raycasting and moving memory, and it has a fast on-chip scratch pad which will have much less latency than writing to external memory (if the PS4's ROPs suffer from ~2x latency of the Xbox One, their ROP throughput will be similar when writing to ESRAM). The leaked PS4 GPU clock is 800mhz and the Xbox One is 853mhz so you've got 12@853 vs 14+4@800. I doubt anyone in this thread knows exactly how that will play out in the long term.

This is where the balanced argument comes from.
smdh0js9n.gif


There are too many wrong things in that post for me to correct .. leaving for work so I'll take a gander later if anyone hasnt addressed your post by then.
 
Don't condescend me. I fully understand the point you are trying to make, and I don't necessarily disagree with it (with regards to launch games).

Yet you spend the rest of your post disagreeing...

And the analogy was obviously a bad idea because you're not able to look at anything outside of the perfectly literal sense. So I'll talk literally, I guess, and see if any of that gets through:

The devs that are working on launch games for these consoles right now don't have time to figure out what makes these machines special. They're working around the clock just trying to get these games ready to ship in November. They were developing software towards a moving target until just months ago. They have bugs to squash, networks to stress test, and in many cases, other platforms to develop and optimize for. If there were large differences between games on the two consoles at this point, that might be revealing. The fact that there aren't large differences is not revealing at all.

Disagree if you like. I can't sum up the reality of a hardware launch any better than that. I've got to go take a dump anyway.
 
Proof is in the pudding. PS3 was supposed to make 360 look like Xbox 1.5.

We all know how that turned out

Why don't you tell us how that turned out because last time I checked the PS3 had some of the most graphically acclaimed games, of any console, this generation.

Also, maybe you didn't get the memo, but these aren't hypothetical projections based on theoretical performance. These are actual performance numbers based on running code. So, yeah, the PS3 didn't make the Xbox look like an Xbox 1.5, but thanks to Microsoft the PS4 is making the Xbox One look like the 360.5.
 
I swear some must enjoy arguing over who's preference is better than actually playing games. Microsoft could not give a shit about you, you are a potential purchase nothing more. The manic nothing to see here cheerleaders versus the fragile stench of the defence force. Take some time to actually play some games like a lot on Gaf and relax, you might enjoy it :-)
 
Top Bottom