Well I'm going to assume these results are based around the engine ubisoft is using.. So these results will vary over numerous engines.. Regardless.. I hope microsoft see's this and realizes they fucked up big time with the hardware in the X1. what the **** were the thinking putting in a GPU like that?.
That being said, the PS3 has double the performance of the 360 last gen, and 99% of the time all games ran better on the 360. I'll still use the X1 as my main console this gen, regardless of raw numbers. The reason the PS4 is selling so well this gen is because of the "hardware advantage".. It's one of the reason's I bought one at launch.. Just 7 months later I sold the thing because I had no games worth playing on the system.
Hardware can only take you so far, software is what matters on consoles. With microsoft's parity clause those raw numbers are nothing but wasted performance that will only ever be seen in some exclusives.. With numbers like that though how did drive club turn out to be such an ugly piece of shit?
@tormentos: Do you see where IT SAY AGENDA ON THE FIRST SCREEN and WHAT VERSION? THE PS4 version thats and hardware not xbox one final Async Compute yes its at 1600 and I cant wait to they start using it more but you to say what you said is ridiculous.
The document is call Efficient usage of compute shaders on XBOX ONE and PS4,not just PS4 the chart compare benchmark for both and the GPU one destroy the xbox one,funny how you were fast to mount into the see the xbox one has a faster CPU like i have been saying,but when is GPU time you start making excuse...lol
Look the XBO CPU can barely do more, 113 vs 98 dancers difference 15 freaking dancers more.
Now look at the GPU 1600 Dancers vs 830 dancers gap 770 Dancer almost 100% gap.
We are talking about compute here,and area where the PS4 was modify beyond the xbox one,it has been know since the first time the PS4 was introduce in February 2013,it has 8 Aces vs the xbox one 2 Aces and 64 commands vs the xbox one 16.
It shows a huge gap even that this game is DX11 based which should port easier to xbox one period.
Take it like a man on compute the PS4 destroy the xbox one,it is no secret it was expected.
They are using PC DirectX 11 for the Xbox One, and then optimized using PSSL for the PS4... So no shit that PS4 performs so much better when they didn't optimize for the Xbox One, using latest tools... They use the DX11 PC version...
I still think PS4 would come out on top, but the gap wouldn't be as big as this.
Are you for real.?
DX11 on the xbox one is DX11X is not the same on PC.
Also it say on page 5 DX11 FOR PC,the coma xbox one PS4 so unless you want to believe that the PS4 has DX11 it is clear they mean PC,but either way porting DX11 from PC to XBO will be easier than to PS4 PSSL.
The xbox one uses DX11X is a special version which has the same things as PC + the gains of DX12.
You can't see what they are doing, can you? They build it with DX11 for PC, and just run it on Xbox One, then they updates it to PS4 software, and run it on PS4. They didn't touch the Xbox One version, they didn't use DX12, just plain DX11.0, and then run it on Xbox One. So no this isn't a valid test, or could you please point out where they are saying they updated the Xbox One version code, from the PC please? Xbox One does indeed use a special DX11 version, but they didn't use that one, they just used PC code, no shit it's running badly then...
The Xbox One should be capable of ~1300 Gflops of compute based on the spec of the GPU. Looking at the result it seems bandwidth limited which suggests they have not used ESRAM. That could be because they did not optimise as you suggest or it could be because other parts of the physics engine are running on the CPU and the need to keep the data coherent which requires them to keep that data in the DDR3 and suffer the bandwidth penalty related to that.
Considering there are multiple options and we are not privy to the details of the test, how it was performed or code that was ran I do not see any evidence that supports your claim. With the exception of optimising for ESRAM I also find it unlikely that it would really cause much of a performance difference anyway if all it is doing is the GPU compute.
Well I'm going to assume these results are based around the engine ubisoft is using.. So these results will vary over numerous engines.. Regardless..
I hope microsoft see's this and realizes they fucked up big time with the hardware in the X1. what the **** were the thinking putting in a GPU like that?.
That being said, the PS3 has double the performance of the 360 last gen, and 99% of the time all games ran better on the 360. I'll still use the X1 as my main console this gen, regardless of raw numbers. The reason the PS4 is selling so well this gen is because of the "hardware advantage".. It's one of the reason's I bought one at launch.. Just 7 months later I sold the thing because I had no games worth playing on the system.
Hardware can only take you so far, software is what matters on consoles. With microsoft's parity clause those raw numbers are nothing but wasted performance that will only ever be seen in some exclusives.. With numbers like that though how did drive club turn out to be such an ugly piece of shit?
Wrong..
The xbox 360 had a superior GPU the PS3 had a superior CPU,the PS3 edge it in specs by a hair,but the PS3 was super hard to code which lead the xbox 360 to have the advantage graphics wise in most multiplatforms.
Yeah because the xbox one has so many games worth playing...lol
Name them and i hope you don't include any unreleased games.
Yeah how that parity clause held up in Shadow Of Mordor.?
How about in Dragon Age.?
Drive club got a 5 while been a graphics beauty,Ryse got even lower a 4 while been a graphics beauty.
@FastRobby said:
You can't see what they are doing, can you? They build it with DX11 for PC, and just run it on Xbox One, then they updates it to PS4 software, and run it on PS4. They didn't touch the Xbox One version, they didn't use DX12, just plain DX11.0, and then run it on Xbox One. So no this isn't a valid test, or could you please point out where they are saying they updated the Xbox One version code, from the PC please? Xbox One does indeed use a special DX11 version, but they didn't use that one, they just used PC code, no shit it's running badly then...
Ubi soft doesn't use DX11 from PC on xbox one you moron,even if they wanted to,the xbox one has its own API and is optimized like DX12,you are just on denial,all developer including UBI are using the latest XBO SDK not DX11 from PC stop making sorry ass excuses.
Page 5 cloth simulation ported to GPU.
For DX11 on PC,xbox one and PS4
It doesn't say DX11 for xbox one unless you want to claim DX11 is also for PS4.
Page 71 Port from HLSL to PSSL on PS4.
This ^^ is a joke It was build for DX and then port to PSSL on PS4,no matter what Porting code from DX to PS4 APi will never ever be as effective as building it from then ground up on PS4,this mean the gap can be actually higher.
The xbox one use DX11X so any game including AC is been my by that api with the latest SDK not DX11 from PC,which basically is the same sh** either way but with some holes.
The Xbox One should be capable of ~1300 Gflops of compute based on the spec of the GPU. Looking at the result it seems bandwidth limited which suggests they have not used ESRAM. That could be because they did not optimise as you suggest or it could be because other parts of the physics engine are running on the CPU and the need to keep the data coherent which requires them to keep that data in the DDR3 and suffer the bandwidth penalty related to that.
Considering there are multiple options and we are not privy to the details of the test, how it was performed or code that was ran I do not see any evidence that supports your claim. With the exception of optimising for ESRAM I also find it unlikely that it would really cause much of a performance difference anyway if all it is doing is the GPU compute.
He is just making excuses this game is DX11 based porting it from PC to xbox one is an ease,compare to porting it to PS4 where no matter what ported code will never run as good as code build from ground up.
But in not place it say it uses DX11 from PC on xbox one,which is a joke because we all know the xbox one uses only its API DX11X which basically is the same as PC but with lower CPU over head and other gains aswell,so porting code from one to another should be a breeze.
I just find incredibly funny that he has to go and claim the test in invalid,because some lame excuse rather than open his eyes and see that no matter what 18 CU will always work better than 12 in this line of GPU,worse the PS4 is admit by sony modify to run compute and graphics at the same time,it has higher number of Aces and command so they can't just see and think for a second let me see the PS4 was always hype for been modify to compute so there is a chance the modifications sony did did pay off..
No they most elaborate an extremely pathetic excuse on why the test is bad,because for them 12 CU can work exactly like 18,but 18 CU can't work like 24 CU which is odd...hahaha
@scottpsfan14 said:
Now don't get excited here. If the PS3 was using just the PPE, it's numbers would be much more like the 360.
But that is the point the PPE wasn't the only resource of Cell it had 7 working SPE 1 reserve to OS the rest could work for games,and they only used 5 if they used 6 it would have beat the xbox one as well.
And the gap vs the 360 CPU would have been instead of 3 times 4 times better.
@tormentos: No excuse I read everything you just happy to see better numbers for your agenda IT SAYS ON THE FIRST PAGE THE PS4 version of the study not ps4 and XBOX one that means they are using ps4 hardware and code PSSL. Like I said its good to see the ps4 almost hundred percent more power that mean we are moving in a great direction in performance and hopefully it show it the games outside of 3rd party, but to say those are the final numbers to go by is ridiculous.
Does that mean the PS3 cpu is better then the PS4s? Devs had said otherwise. Or is this just comparing gpu functions for the processors?
@m3dude1 said:
the ps3 cpu is better for certain flop heavy tasks yes
No the PS3's cpu is not better for normal cpu tasks then PS4. This test is for parallel workloads which works best with many processors like "gpus".The Cell's SPE's were originally designed for parallel workloads aka gpu functions. That is why the PS3's Cell does so well with this test. If they were using the PPE "cpu" portion of the Cell it would be slower then a Pentium 3...
the majority of cpu code isnt flop heavy dummy
Calling someone a dummy is actually quite funny when yourself are one.... Things like physics is parallel in nature and do require more processing power. Heck even doing Ai and npcs require quite abit of processing power based on numbers and advancement of use. In the last few years they have been looking into using gpgpu based Ai making gaming much more dynamic.
wow do you even know wtf a flop is? the majority of code a cpu will crunch through on virtually any game is not flop heavy code and maps TERRIBLY to a gpu
Apparently you dont.... A flop is floating point operations. quite abit of the code for gaming is "flop heavy" which is the reason why many of these games coming out have a 30 fps cap, these jaguar cpu's cant cut it.
Lemmings am cry, even though lemmings been crying for over a year now, this is nothing new and this still won't go through the thick skulls of delusional lemmings
Lmfao no, Xbox One already has forms of low level access already, DirectX12 is barely going to improve anything on the One, your stuck with the meager shoe box
Also lemmings, goodluck having "cloud power" on your 20 mbps connection, real time graphics rendering over cloud beamed to your console is FAR out still, because majority of people dont even have a connection speed faster then 15 MBPS And with pings above 50 ms.
Unless you live in South Korea or so, where the average person has like a 40 mbps connection, US internet infrastructure and the rest of the world is still not up to par yet for any real cloud processing power in real time with very little latency beamed to your console to improve your games graphics
@tormentos: No excuse I read everything you just happy to see better numbers for your agenda IT SAYS ON THE FIRST PAGE THE PS4 version of the study not ps4 and XBOX one that means they are using ps4 hardware and code PSSL.
Like I said its good to see the ps4 almost hundred percent more power that mean we are moving in a great direction in performance and hopefully it show it the games outside of 3rd party, but to say those are the final numbers to go by is ridiculous.
Yes they are excuse first page it says.
""Efficient usage of compute shader on XBOX ONE AND PS4""
^^ First page
The 4th page show what they discuss the part about the PS4 is were they discuss the PS4 version not that the article was about the PS4 version.
""What is this talk about?•
Cloth simulation ported to the GPU•For PC DirectX 11, Xbox One and PS4""
Page 5 ^^^ and your owned..hahahhaaa
The xbox one is there compared and was benchmarked.
Here is the benchmark the CPU is performing better on the xbox one no denying that,the GPU of the PS4 destroying the xbox one GPU.
No you moron skip to page 71 the code was ported from HLSL to PSSL..hahaha
In other words this game is DX11 based,so if anything the game is not accurate for the PS4 since the code is a port not build from ground up like DX11 one is,the test is not PSSL..lol
Those are final number on that engine that is what UBi learn from that engine on XBO,PS4 and PC...lol
@GoldenElementXL said:
So the PS3 CPU benches higher than the PS4's but all we are talking about is what a joke the Xbox One is?
#SystemWars
And here comes the closet lemming ready like always to defend the poor xbox one.]
The PS3 CPU is using SPE those SPE handle GPU task on PS3,which is the reason why Cell totally destroyed the xbox 360 xenon on this task it is highly parallel which this process is,in fact those 5 SPE almost beat the xbox one CPU to,and there was still 1 not used.
The talks about the xbox one is because on the GPU side the PS4 GPU walk over the xbox one GPU,almost doubling it performance wise.
So the PS3 CPU benches higher than the PS4's but all we are talking about is what a joke the Xbox One is?
#SystemWars
Peak Gflop performance of the PS3 has always been high but the problem was extracting it. Specific scenarios, such as this show the PS3 CPU at its best and the Xbox One and PS4 CPU at their worst, that is the whole reason to have GPU compute and you use the CPU to do stuff the CPU is very good at.
Calling someone a dummy is actually quite funny when yourself are one.... Things like physics is parallel in nature and do require more processing power. Heck even doing Ai and npcs require quite abit of processing power based on numbers and advancement of use. In the last few years they have been looking into using gpgpu based Ai making gaming much more dynamic.
wow do you even know wtf a flop is? the majority of code a cpu will crunch through on virtually any game is not flop heavy code and maps TERRIBLY to a gpu
Apparently you dont.... A flop is floating point operations. quite abit of the code for gaming is "flop heavy" which is the reason why many of these games coming out have a 30 fps cap, these jaguar cpu's cant cut it.
Calling someone a dummy is actually quite funny when yourself are one.... Things like physics is parallel in nature and do require more processing power. Heck even doing Ai and npcs require quite abit of processing power based on numbers and advancement of use. In the last few years they have been looking into using gpgpu based Ai making gaming much more dynamic.
wow do you even know wtf a flop is? the majority of code a cpu will crunch through on virtually any game is not flop heavy code and maps TERRIBLY to a gpu
Apparently you dont.... A flop is floating point operations. quite abit of the code for gaming is "flop heavy" which is the reason why many of these games coming out have a 30 fps cap, these jaguar cpu's cant cut it.
only herms can be this dumb, you must be a herm
aww how original.... go cry some more about 30 fps and the parity conspiracy
@tormentos: Fool what is you talking about are you reading the whole thing? they show you the code and how to reduce BW and everything for the ps4 version that was showing code for PSSL only and Dx11 PC is not the same as Xbox either and they didnt use the API or Drivers for the xbox one. As you said yourself they ported from dx11PC to PSSL and they are showing the ps4 version stop being dumb torm. Ubi are porting like this DX11PC >PS4>Xbox one thats what the devs are saying.
@tormentos: Fool what is you talking about are you reading the whole thing? they show you the code and how to reduce BW and everything for the ps4 version that was showing code for PSSL only and Dx11 PC is not the same as Xbox either and they didnt use the API or Drivers for the xbox one. As you said yourself they ported from dx11PC to PSSL and they are showing the ps4 version stop being dumb torm. Ubi are porting like this DX11PC >PS4>Xbox one thats what the devs are saying.
You freaking moron.
1-It was ported from HLSL to PSSL in other words it is a DX11 port to PS4.
2-The xbox one benchmark used the fu**ing xbox one not a PC,and used the xbox one tools not DX11 from PC,that DX11 doesn't work on xbox one the xbox one has its won which is a similar version with some streamline functions.
3-No you idiot they are talking about how they ported the code to PS4 not that the PS4 code is the only one been there,the article is about xbox one PC and PS4 stated in the very first page.
You should drop the whole shit since the PC version isn't even show on the benchmark,just the PS4 xbox one,PS3 and xbox 360,the PS4 is modify for heavy compute and it shows here the PS4 owned the xbox one silly,can wait to see those games pushing compute hard in a few years and how the xbox one will suffer.
The more CU the more compute you can do,there is no way the XBO closes that gap,this is what the PS4 hardware was modify for..lol
@tormentos: @tormentos: because thats what you always start with when you are a multiplat game you start with PC hence using dx11 PC API
Port to the leading platform PS4 hence porting to PSSL from dxPC and show me where they showed the xbox one API and tools please because I show didnt see them? PS4 version because we know xbox dont use dx11PC tools which they was porting from. Hamster on a wheel?
You talking about 400gflops is modifed lol when you go a whole 1.3tf for async compute on the xbox one the only compute heavy game on console right now is Ryse we will see how ps4 will handle heavy compute when bigger games like the Order or Uncharted come out.
and why all the name calling when you really dont want to see me?
@tormentos: @tormentos: because thats what you always start with when you are a multiplat game you start with PC hence using dx11 PC API
Port to the leading platform PS4 hence porting to PSSL from dxPC and show me where they showed the xbox one API and tools please because I show didnt see them? PS4 version because we know xbox dont use dx11PC tools which they was porting from. Hamster on a wheel?
You talking about 400gflops is modifed lol when you go a whole 1.3tf for async compute on the xbox one the only compute heavy game on console right now is Ryse we will see how ps4 will handle heavy compute when bigger games like the Order or Uncharted come out.
and why all the name calling when you really dont want to see me?
the ignorance, twisting and all around excuses as to why the xbone is weaker are hitting an all time low.
@tormentos: @tormentos: because thats what you always start with when you are a multiplat game you start with PC hence using dx11 PC API
Port to the leading platform PS4 hence porting to PSSL from dxPC and show me where they showed the xbox one API and tools please because I show didnt see them? PS4 version because we know xbox dont use dx11PC tools which they was porting from. Hamster on a wheel?
You talking about 400gflops is modifed lol when you go a whole 1.3tf for async compute on the xbox one the only compute heavy game on console right now is Ryse we will see how ps4 will handle heavy compute when bigger games like the Order or Uncharted come out.
and why all the name calling when you really dont want to see me?
No that is what UBI started with,you can build a game from ground up on PS4 without using DX11,the same on PC as well,on xbox one is DX11.x.
Just because they did not show the toolz doesn't mean they are not there HLSL is compatible between DX11 and DX11.x on xbox one,on PS4 it requires porting which is more work not less work.
The xbox one doesn't use DX11 from PC for the 100 times you spinning dolphin,the xbox one use DX11.x and what Ubi did was use DX11.x on xbox one because that is all the xbox one has,there is no other tool period.
You can't use 1.3 TF on the xbox one to compute you moron,what the fu** will run the graphics.?
Another stupid ass post from you now you think the xbox one can use 1.3Tf of compute..hahahaaaaaaaaaa
The real asyno compute is run on the PS4,without trick and work around,the xbox one wasn't modify for compute and if it use 400Gflops for compute it losses them for rendering,that mean they get only 900Glops for rendering while the PS4 would still have 1460Gflops for rendering.
Ryse is a piece of sh** game and enclosed crap which run better and at 1080p on a 7770..hahaha
Infamous use compute you sad fanboy is the reason it has come incredible effects,but but but the PS4 is not doing compute..haahah
Hahaha the xbox one now has 1.3gflosp for compute..ahahahaa that is a good one,i await your usual spin and fix of that mess you just stated there.
If the xbox one uses 4 CU to do compute it is left with 8,i wonder from where in hell you pulled that..lol
@tormentos: I said heavy compute I never said the ps4 wasnt using it.
You moving the goal post again SHOW ME WHERE THEY WAS USING XBOX ONE API porting to ps4 instead of the other way around you just said about that the DX11 dx dont work with XBOX make up your mind lol.
You are crazy on the specs torm THE CONSOLES DONT TRY TO ACHIEVE HEAVY COMPUTE THE SAME WAY.
Yes torm ISS got some nice colors thats it.
It dont matter what you say about Ryse I have played the game hell its the only game I havent traded in one of the best launch games in my book and graphics we already know.
You will not see anything in performance from heavy async compute come out for the xbox one or anything like it until AMD and NVidia release the HBM cards you can bookmark this shit right here if you need to please.
@tormentos: I said heavy compute I never said the ps4 wasnt using it.
You moving the goal post again SHOW ME WHERE THEY WAS USING XBOX ONE API porting to ps4 instead of the other way around you just said about that the DX11 dx dont work with XBOX make up your mind lol.
You are crazy on the specs torm THE CONSOLES DONT TRY TO ACHIEVE HEAVY COMPUTE THE SAME WAY.
Yes torm ISS got some nice colors thats it.
It dont matter what you say about Ryse I have played the game hell its the only game I havent traded in one of the best launch games in my book and graphics we already know.
You will not see anything in performance from heavy async compute come out for the xbox one or anything like it until AMD and NVidia release the HBM cards you can bookmark this shit right here if you need to please.
The only one moving goal post it is you.
Stop spinning i say HLSL ported to PSSL which is what the article state,DX11 and DX11.x on xbox one use the same HLSL language dude.
Infamous use heavy compute you loss the argument.
That launch game is a 60 is mediocre crap you have low standards.
You are joke and trying to imply that you can use 1.3Gflops of power for compute on the xbox one is a joke,you can't..
Basically one more time it is confirmed that you know shit about this.
One more time just to poke you on the eye again..lol
Look the xbox one is there benchmarked by UBI and getting its ass whoop...
I'm glad I stayed out of this one. So much gibberish but hey, it looks like @tormentos is going to win his first debate! I really hope @Tighaman picks his battles better in the future. There is no way anyone with a brain can think that once new GPU's are released next year, the Xbox One's 2 year hardware is going to get better. That makes no sense at all.
And Ryse is not a good game. I know because I own it.
I'm glad I stayed out of this one. So much gibberish but hey, it looks like @tormentos is going to win his first debate! I really hope @Tighaman picks his battles better in the future. There is no way anyone with a brain can think that once new GPU's are released next year, the Xbox One's 2 year hardware is going to get better. That makes no sense at all.
And Ryse is not a good game. I know because I own it.
Actually i have win basically all when it comes to this topic.
The dude is loss claiming the xbox one can use 1.3TF of compute is a joke with what it will run graphics..hahaha
Log in to comment