AnandTech: Xbox One vs. PS4 - Hardware comparison


Recommended Posts

yep you can bet that was a very nice addition to the discussion. anyways, I see that you and a lot of people are going by one little guestimation and so getting your 1.23tflops. and that is guessing that the xbox One is running at the same clock speed that the ps4 is running at which is 1.6Ghz. but a lot people forget that the AMD processor can go 2Ghz. so Microsoft could have a different clock speed that sony.

All we know is that it is an AMD Jaguar, the PS4 could have it clocked to 2GHz also.

Something has been bothering me for a while now but I just couldn't find much answers to go by maybe someone here can help me look into it.

Microsoft quoted that they had 5billion transistors.....but it just doesn't add up:

32md of esRAM is 1.8billion transistors (roughly)

Amd's jaguar I put between 1.2 and 1.7billion (based on on the fact bulldozer has 1.2 transistor count)

1.8+1.7 = 3.5 so what's taking up the rest of the die? There's a whole 1.5billion transistors not counted for.

What is Microsoft hiding? I thought it weird that they would tout there transistor count without giving more details as if they were giving hints.

Bulldozer is just the CPU, Jaguar is CPU and GPU, you are also forgetting the four Move Engines. The PS4 is speculated to have approximately 3 billion transistors, subtract the extra transistors for the better GPU add on ESRAM and Move Engines and theres your 5 billion.

Also thanks to your constant negativism against anything Microsoft ;)

I don?t know if is so so difficult to have both platforms and enjoy them like a true gamer, it is a $$ problem? living room space problem?

Didn't realise being negative wasn't allowed. I go by facts not by what my feelings are, I love talking about technology and am super excited about the next generation of consoles.

I was going to get both, I got a 360 at launch and got rid of it in 2011 when I bought PS3 not long before and realised what I was missing out on.

TV, TV, TV, TV, TV, TV, Sports, Sports, TV, TV, Call a doody, Call a doody, Call a doody, TV, TV, mandatory kinect, all the fuss over wether or not it is going to block used games, all the fuss wether or not you need to be connected to the internet for it to work, doesn't interest me, I don't need a $400-500 TV Box, I use XBMC for all my media consuming needs.

Link to comment
Share on other sites

It should be obvious what you forgot from the equation.

The GPU.

Don't know how you could manage to forget it.

The GPU isn't separate from the processor so the transistor count shouldn't be much different

Link to comment
Share on other sites

It is with Bulldozer. A quad core A8 chip which has Bulldozer core and AMD 3850 is 1.5B transistors. Jaguar has 8 cores and a 7xxx series. The closest GPU for arguments sake is 7770 which has 1.5B transistors.

An 8 Core Bulldozer has 1.2B Transistors add on 1.5B for the 7770, 1.8B for ESRAM and you are talking 4.5B, not sure what the count is on the Move Engines but that is your 5Billion transistors there, no magic extra logic that is going to bridge a massive performance deficit.

Link to comment
Share on other sites

All we know is that it is an AMD Jaguar, the PS4 could have it clocked to 2GHz also.

Bulldozer is just the CPU, Jaguar is CPU and GPU, you are also forgetting the four Move Engines. The PS4 is speculated to have approximately 3 billion transistors, subtract the extra transistors for the better GPU add on ESRAM and Move Engines and theres your 5 billion.

Didn't realise being negative wasn't allowed. I go by facts not by what my feelings are, I love talking about technology and am super excited about the next generation of consoles.

I was going to get both, I got a 360 at launch and got rid of it in 2011 when I bought PS3 not long before and realised what I was missing out on.

TV, TV, TV, TV, TV, TV, Sports, Sports, TV, TV, Call a doody, Call a doody, Call a doody, TV, TV, mandatory kinect, all the fuss over wether or not it is going to block used games, all the fuss wether or not you need to be connected to the internet for it to work, doesn't interest me, I don't need a $400-500 TV Box, I use XBMC for all my media consuming needs.

1) Actually the PS4 can't be clocked at 2ghz also because they already quoted their tflops as 1.84tflops. If if it was a 2ghz chip then the tflops would shoot up to 2.3

Yeah I left out the GPU. At first I didn't think it would add much in the way of transistor count. But the 7790 and 7850 amd cards has 2billion transistors each so 2billion GPU 1.8billion esram 1.2billion CPU that's 5billion on the dot. However this doesn't count the move engines. Unless Microsoft used the extra transistors from the 16 ROPs for the move engines. Seeing that less rops I believe Microsoft should be able to raise the gpus clock speed. (And overclocked 7790 with 16rops is faster than a 7850 with 32rops). Anyways there's a lot about the Xbox one we don't know but we're just guesstimating. The size of the heatsink Microsoft is using is huge I definitely believe they're using higher clock speeds for both CPU and GPU.

Nope you go on rumors too. I've read many of your comments that went all "rumors are facts". And Larry hyrd said that e3 is all about games that there will be TVs on stage but that's the extent in which they'll talk about tv. So yeah you can go check his tweets if you don't believe. So they got out all the tv stuff out the way so now it's going to be all about games.

Link to comment
Share on other sites

Why can't the CPU be clocked at 2GHz, 1.84TF count is for the GPU only.

Xbox GPU isn't a 7790, the 7790 has 896 Cores, it is somewhere between 7770 and 7790 but closer to the former rather than the latter.

Not sure how you figure that an overclocked 7790 (which the Xbox GPU isn't) is faster than 7850. The 7850 has 256-bit memory bus compared to 7790 128-bit, 7850 60% more memory bandwidth, 70% higher pixel rate, twice the number of ROPS. The 7790 is clocked at 1GHz vs 7850 850MHz but is still slower.

OK I go on rumours, but I don't make up stuff like you do about there being some magic extra logic that you are hoping might be there, or that Microsoft might have clocked the CPU/GPU faster than the PS4, I am going from the information that is out there.

I know E3 is about games, I was explaining to Draken why I won't be buying Xbox.

Link to comment
Share on other sites

Why can't the CPU be clocked at 2GHz, 1.84TF count is for the GPU only.

Xbox GPU isn't a 7790, the 7790 has 896 Cores, it is somewhere between 7770 and 7790 but closer to the former rather than the latter.

Not sure how you figure that an overclocked 7790 (which the Xbox GPU isn't) is faster than 7850. The 7850 has 256-bit memory bus compared to 7790 128-bit, 7850 60% more memory bandwidth, 70% higher pixel rate, twice the number of ROPS. The 7790 is clocked at 1GHz vs 7850 850MHz but is still slower.

OK I go on rumours, but I don't make up stuff like you do about there being some magic extra logic that you are hoping might be there, or that Microsoft might have clocked the CPU/GPU faster than the PS4, I am going from the information that is out there.

I know E3 is about games, I was explaining to Draken why I won't be buying Xbox.

ok let me just say this now: I've seen you posted things which gave me the idea that you knew what you were talking about, but now I think I've changed my mind about that. I mean, you don't even know how they reached those teraflops numbers.

how do they reach the tflops? the ps4 has 1152 stream processors (GPU) and is running at 1.6ghz cpu speed: that's 1152x1.6 = 1.843tflops (single precision)

the xbox One has 768 stream processors (GPU) and is rumored to be running at 1.6ghz cpu speed: that's 768x1.6 = 1.23tflops (single precision)

that's how the TFLOPs numbers, the processor speed is accounted for. If I am wrong prove it.

quoting TFLOPs now a days doesn't really say much because it doesnt' take into account other things like GPU frequency, bandwidth, latency, architecture differences. therefore these TFLOPs numbers are more a ballpark figure if anything.

how do I know because I looked at the overclock reviews that's how I know an overclocked 7790 beats a 7850. Plus Sony 7850 is underclocked.

I also think sony might be having heat issues because they're using a lower gpu clock speed from 860mhz to 800mhz.

plus the 7850 uses amd's GCN v1 while the 7790 uses amd's GCN v1.1 (updated architecture remember tflops numbers doesn't take this into account).

the 7850 has a 150w TDP while the 7790 has an 85w TDP

GDDR5 runs much much hotter than DDR3 and uses much more power

The fast esram would take care of any bandwidth bottleneck. Also esram is way less power hungry than the edram used in the 360.

So far it seems the Xbox one should run much cooler and quieter than the ps4 and use way less power. Unless Microsoft changed the estimated clock speeds which would explain the large heatsink and fan.

Link to comment
Share on other sites

ok let me just say this now: I've seen you posted things which gave me the idea that you knew what you were talking about, but now I think I've changed my mind about that. I mean, you don't even know how they reached those teraflops numbers.

how do they reach the tflops? the ps4 has 1152 stream processors (GPU) and is running at 1.6ghz cpu speed: that's 1152x1.6 = 1.843tflops (single precision)

the xbox One has 768 stream processors (GPU) and is rumored to be running at 1.6ghz cpu speed: that's 768x1.6 = 1.23tflops (single precision)

that's how the TFLOPs numbers, the processor speed is accounted for. If I am wrong prove it.

quoting TFLOPs now a days doesn't really say much because it doesnt' take into account other things like GPU frequency, bandwidth, latency, architecture differences. therefore these TFLOPs numbers are more a ballpark figure if anything.

how do I know because I looked at the overclock reviews that's how I know an overclocked 7790 beats a 7850. Plus Sony 7850 is underclocked.

I also think sony might be having heat issues because they're using a lower gpu clock speed from 860mhz to 800mhz.

plus the 7850 uses amd's GCN v1 while the 7790 uses amd's GCN v1.1 (updated architecture remember tflops numbers doesn't take this into account).

the 7850 has a 150w TDP while the 7790 has an 85w TDP

GDDR5 runs much much hotter than DDR3 and uses much more power

The fast esram would take care of any bandwidth bottleneck. Also esram is way less power hungry than the edram used in the 360.

So far it seems the Xbox one should run much cooler and quieter than the ps4 and use way less power. Unless Microsoft changed the estimated clock speeds which would explain the large heatsink and fan.

That's rich, telling me I don't know what I am talking about, you don't even know how you work out TFLOPs number.

For GCN (Graphics Core Next) to calculate TFLOPS first you need to find out the Cores

SIMDS * CU = Stream Processors, or Cores if you like

Cores * Core Clock Speed (stream processors not CPU clock) * 2 (Multiply/Add instructions, MAD, or basically how much work the cores can do per thread) = Gigaflops

PS4

64 * 18 = 1152 Cores

1152 * 800 * 2 = 1843 GFLOPS

1843GF = 1.84TF

Xbox One

64 * 12 = 768 Cores

768 * 800 * 2 = 1228 GFLOPS

1228GF = 1.23TF

We know the Xbox and PS4 are using GCN 1.0 since there has been engineers who worked on both AMD APUs putting the information in the Linked-in profiles and twitter which I think have been removed now. Which is why I said the Xbox GPU is closer to 7770 than 7790.

It's also worth noting that neither console will be using desktop GPUs, they are probably based on mobile versions so comparing TDP is moot. I only said 7770 and 7850 as they are relatively comparative to the GPU in both consoles.

ESRAM helps avoid bottlenecking but I am not convinced that it will totally eliminate it considering 32MB is pretty much GPU framebuffer.

I agree the Xbox One will run quieter and cooler.

Link to comment
Share on other sites

Don't forget as well that Jaguar is based on the 3rd-generation "Steamroller" design. So between that and Piledriver there is probably a slight increase in transistors over Bulldozer that may account for some of the discrepancy.

Link to comment
Share on other sites

how do I know because I looked at the overclock reviews that's how I know an overclocked 7790 beats a 7850. Plus Sony 7850 is underclocked

I'm curious where those reviews you claim show a 7790oc beating a 7850 are, cause i sure can't find one. Of course, it doesn't really matter though, since the one isn't a 7790 anyway.

Link to comment
Share on other sites

That's rich, telling me I don't know what I am talking about, you don't even know how you work out TFLOPs number.

For GCN (Graphics Core Next) to calculate TFLOPS first you need to find out the Cores

SIMDS * CU = Stream Processors, or Cores if you like

Cores * Core Clock Speed (stream processors not CPU clock) * 2 (Multiply/Add instructions, MAD, or basically how much work the cores can do per thread) = Gigaflops

PS4

64 * 18 = 1152 Cores

1152 * 800 * 2 = 1843 GFLOPS

1843GF = 1.84TF

Xbox One

64 * 12 = 768 Cores

768 * 800 * 2 = 1228 GFLOPS

1228GF = 1.23TF

We know the Xbox and PS4 are using GCN 1.0 since there has been engineers who worked on both AMD APUs putting the information in the Linked-in profiles and twitter which I think have been removed now. Which is why I said the Xbox GPU is closer to 7770 than 7790.

It's also worth noting that neither console will be using desktop GPUs, they are probably based on mobile versions so comparing TDP is moot. I only said 7770 and 7850 as they are relatively comparative to the GPU in both consoles.

ESRAM helps avoid bottlenecking but I am not convinced that it will totally eliminate it considering 32MB is pretty much GPU framebuffer.

I agree the Xbox One will run quieter and cooler.

I see you pegged both CPUs at 2ghz correct?

for Microsoft the only thing that has been confirmed so far is the 768 stream processors in the gpu. no clock speeds, no frequencies, no tflops. we're all basing this off rumors, conjectures and leaks. if the xbox one do happen to be what all these rumors and leaks say then so be it. sony did mod the gpu by adding an extra bus lane. I wondering why no one things Microsoft won't mod theirs also.

Link to comment
Share on other sites

Confirmation on the gpu clock speed really isn't needed, tbh. Simply because the clock speeds needed for a 768sp gpu to hit an equal 1.8tflops are way too high for a console.

Link to comment
Share on other sites

As of the Playstation Meeting the PS4 has it clocked at 1.6GHz, Microsoft for all I know could have overclocked it but they are both trying to fit under a thermal envelope, Microsoft doesn't want another RROD situation. I'd be very surprised if either was clocked higher than the base clock that AMD has set.

The only thing confirmed by the word of Microsoft are the 768 Stream Processors and it is a currently theory that the GPU is running at 800MHz as they are trying to fit within a thermal envelope, all of this information was in the leaks of both consoles which has turned out to be 99% correct so far so we know it is GCN 1.0 as AMD don't do an APU with GCN 1.1, there is an APU coming out "this year" that has an 8xxx series GPU in but seeing as these consoles have been in development for years it's unlikely they will have one of the new APUs.

I don't have concrete information but my theory is that:

1) We know the APU is SoC, everything is contained on die, there is a physical limit to how much you can fit into the die which Microsoft has put a weaker GPU in saving space for the very costly ESRAM and Move Engines

2) They are betting on the cloud, they are trying to offset processing so they don't have to have a more powerful GPU

"The box will pop on and come to your home page or wherever you were last. In order to do that in an efficient way, you have to architect all of that into the box up front. A lot of it is in the SoC," Holmdahl says.

That SoC contains both the CPU and GPU, as well as embedded ESRAM; the first two components are based on an AMD design, and custom-built into an SoC with that embedded memory.

Your theory that the ESRAM is external to the die is incorrect.

The console runs in multiple power states, which means it runs in a low-wattage setting when not in use. (Microsoft wouldn't give us specifics other than to say, "The system is designed for an SoC up to about 100W, but will vary on the scenario.")

http://www.engadget....an-inside-look/

Straight from the horses mouth, they are trying to fit within a 100W TDP envelope which is why I believe neither the CPU or GPU will be running any faster than PS4.

Link to comment
Share on other sites

I'm curious where those reviews you claim show a 7790oc beating a 7850 are, cause i sure can't find one. Of course, it doesn't really matter though, since the one isn't a 7790 anyway.

I saw it over at xbitlabs. granted the OCed 7790 didn't sweep the floor but it did win some, mostly with AA off.

it won in 3dmark 2011.

it was close in 3dmark 2013, but they said it was hampered by it's low onboard memory.

it was also close in unigine valley bench

not by much but also in certain games like resident evil 6, stalker, F1 2012, borderlands 2 and some others.

when all was said and done, the two cards averaged out at 48.3 and 49.8 with AA off. but with AA4x no it's 27 to 32

anyways, the 7790 OC lost more than it won. but my point was it still beat it.

also keep in mind that the ps4's gpu is underclocked down to 800mhz so it won't give the same results as the test results show on the charts it would be a bit less.

http://www.xbitlabs.com/articles/graphics/display/radeon-hd-7790-geforce-gtx-650ti-boost.html

Link to comment
Share on other sites

The PS4 GPU isn't a 7850, it is just the closest comparison we have and the Xbox GPU isn't anywhere near 7790 as it's a GCN1.0 not 1.1. The rumours are that the PS4 GPU is 7970M with 2 less CU.

Link to comment
Share on other sites

The PS4 GPU isn't a 7850, it is just the closest comparison we have and the Xbox GPU isn't anywhere near 7790 as it's a GCN1.0 not 1.1. The rumours are that the PS4 GPU is 7970M with 2 less CU.

if you're going to say the xbox one is more closer to a 7790 then I'm sure you know that the 7790 runs at 1000mhz (1ghz) clock speed. therefore going by your formula the xbox would be 768 x 1000 x 2 = 1.54tflops and not 1.23

Link to comment
Share on other sites

What?

You are the one saying Xbox is closer to 7790 when in reality it isn't. It's closer to 7770 but it is just a comparison, the APUs use mobility versions of the 7xxx series.

Also it's not 'my' formula, it is 'the' formula. I like how you call me out and basically tell me I don't know what I'm talking about and you ignore most of what I post to try and fit your own theories.

I'm done here.

Link to comment
Share on other sites

What?

You are the one saying Xbox is closer to 7790 when in reality it isn't. It's closer to 7770 but it is just a comparison, the APUs use mobility versions of the 7xxx series.

Also it's not 'my' formula, it is 'the' formula. I like how you call me out and basically tell me I don't know what I'm talking about and you ignore most of what I post to try and fit your own theories.

I'm done here.

you're saying that I ignore most of what you said to fit my theories? while you're not doing the same? I'm puzzled as in how you get to the point to know that the xbox One's gpu is more closer to a 7770 for all we know it could be a chopped down 7950.

And when have I said that the esram was external. I've been adding up transistor count so how could I say the esram was external. Also anandtech estimates the transistor count for the esram was 1.6billion. I thought it was 1.8.

Link to comment
Share on other sites

You do ignore what I say, it's clear you don't have a clue when you think TFLOPs are calculated by timesing the core count by CPU clock speed.

It doesn't matter what it is chopped down from, all we need to know is how many cores, rops, tmus, clock speed it has and we can deduce a relative level of performance from that whether it is a cut down 7990 or a built up 7350.

I know for a fact that it's not a GCN 1.1 GPU as there are no mobile 1.1s.

From the specs of both machines we can do a relative comparison between the 7770 and 7850 as they are the closest GPUs and have a similar gap in specs as the PS4 and Xbox One.

Also Anand from Anandtech website says that it would require a 66% jump in TDP to run the Jaguar at 2GHz instead of 1.6GHz so that disproves another one of your "theories"

Link to comment
Share on other sites

You do ignore what I say, it's clear you don't have a clue when you think TFLOPs are calculated by timesing the core count by CPU clock speed.

It doesn't matter what it is chopped down from, all we need to know is how many cores, rops, tmus, clock speed it has and we can deduce a relative level of performance from that whether it is a cut down 7990 or a built up 7350.

I know for a fact that it's not a GCN 1.1 GPU as there are no mobile 1.1s.

From the specs of both machines we can do a relative comparison between the 7770 and 7850 as they are the closest GPUs and have a similar gap in specs as the PS4 and Xbox One.

Also Anand from Anandtech website says that it would require a 66% jump in TDP to run the Jaguar at 2GHz instead of 1.6GHz so that disproves another one of your "theories"

With the gcn 1.1 I was going by what anand wrote about the 7790.

also yeah it could be and modified 7350 who knows, not I, not you, not anyone but Microsoft. again, most what we're doing are just pure guestimations .

About the jump in TDP AGAIN maybe that explains the huge heatsink and fan Microsoft put in the x1 that I've been mentioning. throughout their system Microsoft uses low powered parts yet they have a really big heatsink? sounds to me they increased something be it cpu clock speeds or gpu clock speeds but something got increased.

I came across this online:

"The PS4's GPU can read/write to it's 8GB of GDDR5 at 176 GB/sec only

The ONES GPU can read/write it's 8GB of DDR3 at 68 GB/Sec

The ONES GPU can read/write it's 32MB SRam frame buffer at 102GB/sec

The ONES GPU can read/write to its 4 Move engines at 25.6 GB/Sec

The ONE GPU can read/write to all of the above simultaneously

Assuming the PS4 wants games to look as good as XO games with no compromises, they will need to allocate 102GB/Sec to the framebuffer (1080p game is 16mb)

This leaves us with the remaining bandwidth

The PS4's GPU can read/write to it's 8GB of GDDR5 at 74 GB/sec

The ONES GPU can read/write it's 8GB of DDR3 at 68 GB/Sec

The ONES GPU can read/write to its 4 Move engines at 25.6 GB/Sec

Now if this is the case then bandwidth wise the Xbox technically has better throughput. Which is why Microsoft mentioned 200GBs because all these memory pools can all be accessed simultaneously.

When microsoft quoted 200 Anand wrote that "this is not how it works in the real world" but I think he was looking at it in the traditional sense as in oneway throughput. I need to ask him about this and how plausible it is.

He did say the Xbox one has a more complicated setup while Sony went with a more simpler "increase the memory lane and call it a day"

Link to comment
Share on other sites

Can't be bothered discussing it any more, people have already erroneously made their mind up, coming up with all sorts of fantastical theories about magic circuitry that Microsoft might have put in the machine.

At the end of the day PS4 is 50% faster and Microsoft are betting on the cloud to help them bridge the gap and I think it's a huge mistake.

Link to comment
Share on other sites

Can't be bothered discussing it any more, people have already erroneously made their mind up, coming up with all sorts of fantastical theories about magic circuitry that Microsoft might have put in the machine.

At the end of the day PS4 is 50% faster and Microsoft are betting on the cloud to help them bridge the gap and I think it's a huge mistake.

I actually think Microsoft is just trying to innovate and find a way to utilize the cloud in an effort to actually make it useable. They've already said what they may use it for and that is not, and cannot ever be real time rendering. Not only that, the difference in GPU power does not require the use of the cloud to make up the difference. You're just being fantastic with the whole the XBOX will perform at 50% of the effective performance of the PS4 so they need cloud processing to make up for it. That's actually a hilarious position to take. And I'm not even that big on Xbox One. I'll probably go from my current Xbox to a PS3 with Xbox controller and skip next gen unless something dramatic happens.

Get yourself a PC with identical GPUs one with GDDR3 and one with GDDR5 and run benchmarks. See if the performance differences are anywhere near the theoretical differences. The PS4 will have an edge, but nowhere near leapfrog or 50% overall performance gain. In fact, have multiple OS make may many functions feel faster on Xbox One.

Remember how technically superior the Cell is in PS3, yet how troublesome XMB functions during games was?

Link to comment
Share on other sites

Get yourself a PC with identical GPUs one with GDDR3 and one with GDDR5 and run benchmarks. See if the performance differences are anywhere near the theoretical differences. The PS4 will have an edge, but nowhere near leapfrog or 50% overall performance gain. In fact, have multiple OS make may many functions feel faster on Xbox One.

It's not just the difference in memory bandwidth. The Xbox One GPU has 384 less cores, less TMUs, half the ROPS, half the memory bus width. If we do a very unscientific comparison with relative GPUs you are looking at 7750/7770 vs 7850 which at 1080P with graphics on full a good 25FPS~ difference between Xbox One and PS4.

Obviously some of the comparisons in this image are off as they don't exactly match the specs but for arguments sake 7850 vs 7770:

post-1442-0-72120800-1370158857.jpg

Link to comment
Share on other sites

It's not just the difference in memory bandwidth. The Xbox One GPU has 384 less cores, less TMUs, half the ROPS, half the memory bus width. If we do a very unscientific comparison with relative GPUs you are looking at 7750/7770 vs 7850 which at 1080P with graphics on full a good 25FPS~ difference between Xbox One and PS4.

Obviously some of the comparisons in this image are off as they don't exactly match the specs but for arguments sake 7850 vs 7770:

So you're still believing your theory that the xO has a 7770/7790 graphics core huh.

But do you think a 7790 can push 1080p at 60fps from a demanding game like forza?

About the memory bandwidth I think I've already proven that the PS4 doesn't have an edge here and that the Xbox One seems to be the one with the edge (waiting on anand 's reply.

Link to comment
Share on other sites

Can't be bothered discussing it any more, people have already erroneously made their mind up, coming up with all sorts of fantastical theories about magic circuitry that Microsoft might have put in the machine.

At the end of the day PS4 is 50% faster and Microsoft are betting on the cloud to help them bridge the gap and I think it's a huge mistake.

So you admit that all you've been doing is trying to convince others from buying the Xbox One then? Well whatever floats your boat.

Since the is 50% faster then I'm guessing we should see exceptionally beautiful and performing games when compared to the Xbox One right? I'm talking about 1st party not 3rd party games.

From what I've seen of the first party games on ps4 so far isn't fallowing that trajectory of being "50% faster"

Link to comment
Share on other sites

So you're still believing your theory that the xO has a 7770/7790 graphics core huh.

But do you think a 7790 can push 1080p at 60fps from a demanding game like forza?

About the memory bandwidth I think I've already proven that the PS4 doesn't have an edge here and that the Xbox One seems to be the one with the edge (waiting on anand 's reply.

I never once said the xO will have 7790, 7770 is a more realistic comparison but in reality it will be somewhere inbetween as the 7770 has less cores than the xO GPU. For comparisons sake the gap between 7770 and 7850 is going to be similar to the gap between xO and PS4 which is 25FPS+ at 1080p.

Not sure what you are trying to say with your memory bandwidth post, memory bandwidth just means how much data you can pass from system memory to CPU/GPU, the xO is still constrained by its weak GPU and it's not as simple as just adding all the memory bandwidth together just like you can't say that it has a 12.8GHz CPU just because it has 8x 1.6GHz cores.

Mark Cerny the Chief Architect of PS4 did consider a 128-Bit memory bus and EDRAM for PS4 but realised that it would be too complex for developers to take advantage of hence why PS4 has a 256-bit bus with GDDR5.

The PS4 has an additional 20GB/s bus that bypasses the GPU cache if you wanted to do a ###### for tat erroneous adding up of memory bandwidth the xO doesn't have any advantage.

So you admit that all you've been doing is trying to convince others from buying the Xbox One then? Well whatever floats your boat.

Since the is 50% faster then I'm guessing we should see exceptionally beautiful and performing games when compared to the Xbox One right? I'm talking about 1st party not 3rd party games.

From what I've seen of the first party games on ps4 so far isn't fallowing that trajectory of being "50% faster"

So you've seen a couple of pre-alpha PS4 games, no actual ingame footage (as they were running on PCs) and all the pre-rendered crap at the xO reveal and can say with confidence there won't be a gap in performance?

Cerny gave the PS4 an extra bus to bypass the cache, he has enhanced the L2 cache so that graphics and compute threads can be run in parallel and he has increased the compute command "sources" from 2 to 64. As well as the extra 384 cores, double the memory bus width, double the rops, more tmus, no amount of memory bandwidth or magic extra logic is going to bridge the gap.

I don't give a toss if someone buys Xbox One or not I am just making sure people have an informed decision, I don't pull random theories from the air like yourself to desperately try and justify your decision to get an Xbox One.

Link to comment
Share on other sites

This topic is now closed to further replies.
  • Recently Browsing   0 members

    • No registered users viewing this page.