By using this site, you agree to our Privacy Policy and our Terms of Use. Close

Forums - Gaming Discussion - Will the PS4 represent 50% better performance than the X1

ethomaz said:

DirtyP2002 said:

The system floating point performance of the PS3 is twice the system floating point performance of the Xbox 360. That is all I am saying.

First GPU.

RSX is a weak variation of GeForce 7800 GTX... the 7800 GTX have 165 GLOPS performance... you can say RSX run a little faster (450Mhz to 550Mhz) but in fact the RSX is a weak variantion and not strong one... in any case the RSX have at maximum 200GLOPS.

Xenos is a unified arch with 48 vec5 running at 500Mhz... so easy that I will give you the full number directly: 240GFLOPS.

RSX: 165-200 GLOPS
Xenos: 240 GLOPS

Yeah... the GPU differnce is that big.

CPU.

Xenon is a PPC based arch... the VMX units (like the SSE units in Intel) in a 3-core PPC runnin at 3.2Ghz can archive at maximum 80 GFLOPS.

Cell it is the difference here...each SPU can archive 25.6 GFLOPs... PS3 have 7 of them on... so 179,2 GLOPS (I will use 180 GFLOPS).

Cell: 180 GFLOPS
Xenon: 80 GFLOPS

Overall

PS3: 345-380 GFLOPS (the articles in Intenet shows PS3 with 400 GLOPS because they use the FLOPS of the full enable Cell with 8 SPUs)
360: 320 GLOPS

In the best case schenario the PS3 is 20% more powerful than 360... the best case is how much you used the power of Cell in the right way.

In FLOPS the difference is that but remember it's ways hard to use these FLOPS from Cell that makes the diffence be less than what the theorical shows... and 360 have a better arch for memory so even more easy to use memory in 360 than PS3.

The use of Cell for gaphics is a hell of difficult... the 360 have a easy use of 240 GFLOPS for graphics here.

Xbone x PS4

Exactly the same CPU so I won't make the maths here... GPU 50% more powerful for PS4... we never saw a generation with this gap in power between Microsoft and Sony (we saw a even bigger with Nintendo x Sony/Microsoft).

Good work, but you are forgetting a few bus bottlenecks and some other problems.  My calculations put the Xbox 360 and PS3 are more like 5% of each other's power.  (Remember they were both mostly developed at the same place, at the same time at IBM.)

“We never saw a generation with this gap in power between Microsoft and Sony”

Remember the Xbox was more than 100% more powerful than the PS2.  And outside of a few exclusives, how much difference would you say you saw between the two?

Being 100% more powerful is when you will just start to notice.  We will really see the difference compared to current gen’s as the next ones are 8-10 time more powerful than current (X360 PS3.)

 



 

Really not sure I see any point of Consol over PC's since Kinect, Wii and other alternative ways to play have been abandoned. 

Top 50 'most fun' game list coming soon!

 

Tell me a funny joke!

Around the Network
Zappykins said:

While I will confirm Sony has the better GPU, they messed up again.  GDDR5 RAM is great for graphics, it should make nice pretty pictures – But unfortunately, it comes at a cost - huge latency.

I'm not talking about twice as bad, or even 5 times, but 7-10 times as much latency as DDR3 memory.  It will cripple all the other advantages they put in it – like the wider bus width, speed, etc.  It’s really a shame.  It will effect everything thing that isn’t just a picture (AI, Game play, etc.)

 

 

No. GDDR5 does not have massive latency hit compared to DDR3, please stop parroting misinformation.
Go look up the datasheets of the actual chips and you will soon realise latency in RAM has always stayed at about 10ns, infact it's been around that level since DDR1.

However, since datarates on Ram have been increasing over the years so has the latency in clock cycles, but the absolute latency has been relatively static.

Grab some DDR3 1600mhz memory, that's 800mhz IO, which has a typical CAS latency of 8, that means it has a latency of 10ns.
Grab some DDR2 800mhz memory, that's 400mhz IO, which has a typical CAS latency of 4, this is also 10ns.

Now with GDDR5 the data rates are 4x faster than the IO clock instead of 2x, I.E. 5ghz GDDR5 is 1.25ghz x4 and would have a CAS Latency of 15.
15/(1.25 GHz) = 12 ns

So the latency of GDDR5 is only 20% higher than DDR3.



--::{PC Gaming Master Race}::--

I doubt there will be a significant difference. It will just be easier for devs with the ps4.



Oh there will be a noticeable difference, just as multiplats usually ran better on the 360, they'll now run better on the PS4. Situations like Skyrim on PS3 will now be Xbox one situations.

So things like games running below 1080p, games not hitting 60fps, games having low rez effects or reduced effects, games having little or no Anti-aliasing, all this will be an issue on the Xbox one just as it was on the PS3, except that it will be more pronounced.

Keep in mind that the PS3 and 360 were even enough in power, it just took more effort on the PS3 because it's architecture, so while true that the 360 had a better GPU it's CPU wasn't all that, but PS3 CPU, the CELL, could more than enough compensate for it's weaker GPU. Now with the PS4 and Xbox they are for all intent and purposes running the same CPU and GPU tech from AMD, the key difference is memory setup and it's repercussions, mainly that the xbox one SOC sacrificed 6CU in order to embed ESRAM, dropping it's GPU performance potentially from 1.8TF to just 1.2TF....which is a sizable difference in a closed box like the consoles are.

The Xbox one is now left with a weaker/less capable GPU with no other hardware to compensate for it, but one could argue that maybe the ESRAM, like CELL could some how compensate for the GPU weakness, now this could be true had the ESRAM had some crazy bandwidth like 256GB/s thus enabling free AA....but the reality is the ESRAM only does 102GB/s which is less than the PS4 176GB/s, sure if a developer managed to work enough magic to use both the DDR3 and ESRAM combined bandwidth it'll only be 170GB/s which is still slower than PS4 176GB/s...so which ever game hits a bandwidth limit on the xbox one causing lower frame rates, will simply run better on the PS4.

Suffice to say, If Microsoft had a hardware advantage they would be touting it right now, the fact they have yet to release official specs at all is very telling....in all honesty the only advantage the xbox one currently has is the advanced kinect sensor...it's also the main reason I think why the Xbox one cast $499.

So the motto use to be, wanna play multiplats, get a 360, will now be, wanna play multiplats get a PS4, as for 1st party exclusives, as always get the console you prefer.



Pemalite said:
Measuring in flops is pointless, it's not an accurate representation of performance.

For starters, the Cell will *only* achieve anywhere near it's theoretical performance with linear equations, otherwise it's orders of magnitudes slower.

The Xbox 360 and PS3 are very different beasts from an architectural perspective, different memory set-ups, vastly different GPU's, different CPU's etc'.
The Xbox One and PS4 on the other hand are almost identical from an architectural perspective, minus the differences in the memory systems.

Whichever way you cut the cake though, with the Xbox One's 50% slower GPU... There is going to be differences in image quality later in the machines life when developers push the hardware, think: Less Geometry, Texture resolution due to less available Ram, Particle counts severely scaled back... That sort of thing.

This is only true if the GPU tech and manufacturers are different, AMD flops and Nvidia flops aren't 1:1, however, both the PS4 and xbox one are using the same manufacturer, AMD,  with effectivly the same GPU tech, the difference being that the PS4 has 18CU while the Xbox one has 12CU....so in this situation using FLOPS as measurment is far from pointless, rather it makes it easier to gadge the difference with a higher accuracy of true performance.



Around the Network
jake_the_fake1 said:

This is only true if the GPU tech and manufacturers are different, AMD flops and Nvidia flops aren't 1:1, however, both the PS4 and xbox one are using the same manufacturer, AMD,  with effectivly the same GPU tech, the difference being that the PS4 has 18CU while the Xbox one has 12CU....so in this situation using FLOPS as measurment is far from pointless, rather it makes it easier to gadge the difference with a higher accuracy of true performance.


No, it is pointless, please read my post within it's entire context. :)

An example would be:
A GPU can have 10,000 Teraflops, but if a GPU with 1 Teraflop has say... 10x the Geometry performance, then graphically, it can pull ahead, because the world won't be all flat boring surfaces. :)

Besides, we all know that the Xbox One has 50% less GCN pipelines, but the question begs on what other parts of the GPU are in actual fact exactly the same (Possibly better? Who knows.) as the PS4, we probably won't know untill release.



--::{PC Gaming Master Race}::--

Pemalite said:
jake_the_fake1 said:

This is only true if the GPU tech and manufacturers are different, AMD flops and Nvidia flops aren't 1:1, however, both the PS4 and xbox one are using the same manufacturer, AMD,  with effectivly the same GPU tech, the difference being that the PS4 has 18CU while the Xbox one has 12CU....so in this situation using FLOPS as measurment is far from pointless, rather it makes it easier to gadge the difference with a higher accuracy of true performance.


No, it is pointless, please read my post within it's entire context. :)

An example would be:
A GPU can have 10,000 Teraflops, but if a GPU with 1 Teraflop has say... 10x the Geometry performance, then graphically, it can pull ahead, because the world won't be all flat boring surfaces. :)

Besides, we all know that the Xbox One has 50% less GPU hardware, but the question begs on what other parts of the GPU are in actual fact exactly the same as the PS4, we probably won't know untill release.

 

Your example is not applicable to the PS4/Xbox one because they both have almost identical GPU tech.

Now if the Xbox one has a ray tracing unit, then yeah even with a weaker GPU the lighting the Xbox one would have would crush whatever the PS4 could do, then your example would be applicable.

So far though the specs which have been revealed were correct for the PS4 so we can currently take it that the xbox one rumoured specs are true as well...obviously taken with a grain of salt, but again there has been nothing to say that the Xbox one has anything special that all of a sudden change the playing field, if there was Microsoft would be talking about it especially with the PS4 specs officially being revealed....funny that Microsoft is being so vague, almost Nintendo like….

Your right though, once these boxes are released, they will get gutted and will find out exactly what's underneath the xbox one hood....I'm just saying that there probably won't be anything new that we don't already know, the gutting of the box will just confirm what’s been rumoured.



jake_the_fake1 said:
Pemalite said:
jake_the_fake1 said:

This is only true if the GPU tech and manufacturers are different, AMD flops and Nvidia flops aren't 1:1, however, both the PS4 and xbox one are using the same manufacturer, AMD,  with effectivly the same GPU tech, the difference being that the PS4 has 18CU while the Xbox one has 12CU....so in this situation using FLOPS as measurment is far from pointless, rather it makes it easier to gadge the difference with a higher accuracy of true performance.


No, it is pointless, please read my post within it's entire context. :)

An example would be:
A GPU can have 10,000 Teraflops, but if a GPU with 1 Teraflop has say... 10x the Geometry performance, then graphically, it can pull ahead, because the world won't be all flat boring surfaces. :)

Besides, we all know that the Xbox One has 50% less GPU hardware, but the question begs on what other parts of the GPU are in actual fact exactly the same as the PS4, we probably won't know untill release.

 

 

Your example is not applicable to the PS4/Xbox one because they both have almost identical GPU tech.

Now if the Xbox one has a ray tracing unit, then yeah even with a weaker GPU the lighting the Xbox one would have would crush whatever the PS4 could do, then your example would be applicable.

So far though the specs which have been revealed were correct for the PS4 so we can currently take it that the xbox one rumoured specs are true as well...obviously taken with a grain of salt, but again there has been nothing to say that the Xbox one has anything special that all of a sudden change the playing field, if there was Microsoft would be talking about it especially with the PS4 specs officially being revealed....funny that Microsoft is being so vague, almost Nintendo like….

Your right though, once these boxes are released, they will get gutted and will find out exactly what's underneath the xbox one hood....I'm just saying that there probably won't be anything new that we don't already know, the gutting of the box will just confirm what’s been rumoured.



Well that's the thing, things might not be as similar as you think. The X1 has four LZ77 Move Engines built into the GPU that compresses and decompresses in the background. Something that the PS4 GPU obviously does not.  The audio chip in the X1 is supposedly a powerhouse as well (presumably to handle Kinects vastly improved voice recognition, amongst other things). This might end up like the PS3/360, one system "appears" to be more powerful but in the end they end up much more comparable because of other factors beyond the Teraflop count.



MaulerX said:
jake_the_fake1 said:
Pemalite said:
jake_the_fake1 said:

This is only true if the GPU tech and manufacturers are different, AMD flops and Nvidia flops aren't 1:1, however, both the PS4 and xbox one are using the same manufacturer, AMD,  with effectivly the same GPU tech, the difference being that the PS4 has 18CU while the Xbox one has 12CU....so in this situation using FLOPS as measurment is far from pointless, rather it makes it easier to gadge the difference with a higher accuracy of true performance.


No, it is pointless, please read my post within it's entire context. :)

An example would be:
A GPU can have 10,000 Teraflops, but if a GPU with 1 Teraflop has say... 10x the Geometry performance, then graphically, it can pull ahead, because the world won't be all flat boring surfaces. :)

Besides, we all know that the Xbox One has 50% less GPU hardware, but the question begs on what other parts of the GPU are in actual fact exactly the same as the PS4, we probably won't know untill release.

 

 

 

Your example is not applicable to the PS4/Xbox one because they both have almost identical GPU tech.

Now if the Xbox one has a ray tracing unit, then yeah even with a weaker GPU the lighting the Xbox one would have would crush whatever the PS4 could do, then your example would be applicable.

So far though the specs which have been revealed were correct for the PS4 so we can currently take it that the xbox one rumoured specs are true as well...obviously taken with a grain of salt, but again there has been nothing to say that the Xbox one has anything special that all of a sudden change the playing field, if there was Microsoft would be talking about it especially with the PS4 specs officially being revealed....funny that Microsoft is being so vague, almost Nintendo like….

Your right though, once these boxes are released, they will get gutted and will find out exactly what's underneath the xbox one hood....I'm just saying that there probably won't be anything new that we don't already know, the gutting of the box will just confirm what’s been rumoured.

 



Well that's the thing, things might not be as similar as you think. The X1 has four LZ77 Move Engines built into the GPU that compresses and decompresses in the background. Something that the PS4 GPU obviously does not.  The audio chip in the X1 is supposedly a powerhouse as well (presumably to handle Kinects vastly improved voice recognition, amongst other things). This might end up like the PS3/360, one system "appears" to be more powerful but in the end they end up much more comparable because of other factors beyond the Teraflop count.

 

The move engines are nothing more than part of the DMA setup, all they've added is some extra compression units, to say that it's something over the PS4 when the PS4 also uses DMA and also have compression units is disingenuous. To be clear, the DMA on the Xbox one has been modified enough in order to keep the ESRAM full and the GPU fed, something the PS4 doens't need to worry about since it has no ESRAM, in other words the Move Engine are nothing special, their just there along with ESRAM to mitigate the low bandwidth of DDR3 memory. The PS4 simply has 8GB of fast ram, 176GB/s, accessible by both the GPU/CPU and this is why neither ESRAM or Move engine are required, 2 approaches to achieve the same goal, the PS4 just achieves the goal better with less headaches.

As for SHAPE, the sound chip, yeah it's a power audio processor from what I understand, however, like you said it'll most likely there to handle kinect related audio processing. the PS4 also has dedicated sound hardware, but since it's not doing kinect like stuff, it's not required but again it's not to be taken as an OMG xbox one has mo powa!

Just keep in mind, Microsoft like Sony love to tout their power or advantage, Sony have officially reviled their system specs, Microsoft has not, all they've been talking about is the power of the 'Cloud' while leaving the system specs vague....kinda strange considering they touted the power of the 360 in the past, and their still touting power now, they’re just keeping quiet and vague about the power of the Xbox one, to me this is telling as Microsoft would not be one to keep silent if they indeed had better specs, but currently it seems that their being silent because indeed their box is weaker than the PS4, hence, the emphasis on the cloud.



5GB GDDR 3 Ram Vs. 8GB GDDR5 Ram..
yes in the long run you will see a difference.