Support NeoGAF

[DForce]

Fine. You do you. But you're kind of agitated based on things I'm not actually doing. If you feel the proof is so clear-cut, pull up my quotes where you think I've engaged in these things you are alleging.

I never said XSX was full RDNA2, in fact I've said it's also custom RDNA2 several times in the past, as well. So if I've said PS5 is not "full" RDNA2 then what I've probably actually said is that it's custom RDNA2, similar to XSX. The question is how the systems have customized their feature sets, and that is a hard question to answer because even AMD are cheeky when it comes to what RDNA2's full feature set is.

From the sounds of it, you maybe didn't watch the NXGamer video as intended, because I don't think that's what they were saying. They were basically touching up on architectural improvements in the RDNA2 GPUs which would help more fully utilize the GPU hardware to reach ever closer to the theoretical performance numbers. Like I said, there won't be any real-world use cases where PS5 actually reaches 10.275, and there won't be any real-world use cases where XSX actually reaches 12.147. But thanks to system designs and architectural gains both systems aught to reach much closer to their theoretical peaks under absolute/highest utilization than their predecessors.

I said MS's numbers are sustained because MS literally used the word "sustained" when stating them. And the reason I take them on their word on that is because the XSX is also being designed with use in server markets, primarily Azure server racks IIRC. Sustained performance is a necessity there, hence why they probably mentioned it (tho tbf, they also probably mentioned it as a cheeky jab against Sony's variable frequency). There's no agenda in stating they've claimed sustained and Sony haven't because Sony haven't literally came out and said their numbers are sustained on the SSD.

Now, I can afford them benefit of the doubt and say they likely are, but I'm also considering the power draw a SSD of that level has to put on the system in terms of potential strain, and the fact that will pretty much factor into the variable frequency. Very logical things to think about, considering Cerny stressed the importance of their variable frequency strategy and what needed to be done to achieve it.

So, I feel I've hopefully explained myself satisfactorily. If for some reason you're still not satisfied with my rationale, or seem to somehow think I'm out to belittle/downplay PS5 or Sony or Mark Cerny out of spite, I can't help you. Because you're literally wrong; I've been critical (in a constructively optimistic way) on design and other aspects on both systems, and will continue to do so.

I don't know how I can make my post easier for you to understand. I watched NX's video several times. Why don't you take the time to READ my post.

I never said XSX was full RDNA2, in fact I've said it's also custom RDNA2 several times in the past, as well.

I never said YOU did.

See, this is why you need READ my post. I made a list of accusations by fanboys who are attacking the PlayStation 5.

Like I said, there won't be any real-world use cases where PS5 actually reaches 10.275, and there won't be any real-world use cases where XSX actually reaches 12.147. But thanks to system designs and architectural gains both systems aught to reach much closer to their theoretical peaks under absolute/highest utilization than their predecessors.

And?
I'm just over here shaking my head. You're missing the point. People are saying it's a 9.2TF console and not a 10.2TF. They were not referring to real world cases, they were talking about the actual TF count. This means if anyone says PS5 is a 10.2TF console, they would say, "No, it's really 9.2TF".


If you're not going to read my post, then dont waste my time, dude.

 

[Panajev2001a]

I agree that 'most' and 'majority' mean pretty much the same thing, unfortunately that's not what we were discussing.

Just to be clear, you posted 'vast majority' which is a different thing altogether. 51% of the time could be 'most', 'vast majority of the time' suggests near constant speeds.

I simply asked if you could link to your source, given the quote, should I assume that you have no source?

I think the PS5 will be at peak performance pretty much all of the time but it might be best to refrain from making up your own Cerny quotes.

Refrain? I did not state anything beyond what he stated in the presentation or the follow up DF article/interview with Leadbetter, but sure let’s play semantics on majority, vast majority, and most of the time. I would ask you to show me where I said that it was an actual quote word by word and where in the Road to PS5 or the follow up DF interview anything is said about clockspeed to indicate otherwise... sue me for paraphrasing I guess .

 

[Panajev2001a]

Care to correct me then?
I mean did I make any mistake with these calculations?
Are they wrong?
Sure the compression rates are assumptions I already said that.


Just so you know based on my example:
7.87 = 100%
5.99 = 76%
Delta = 24% aka XSX is 24% slower then PS5.

Please note that while xsx has nearly 6.0GB/s compressed data in this case the PS5 has only 7.9GB/s which is still not what Cerny said.
I mean this is simple math you can probably tell me if I've made any mistakes.

How do you get to 5.99/6.0 GB/s again, is this the made up 2-3x multiplier or taking the theoretical peak of the BCPack decompressor? If you do that I will take 22 GB/s for the Kraken one in the PS5 please .

 

[Deto]

Care to correct me then?
I mean did I make any mistake with these calculations?
Are they wrong?
Sure the compression rates are assumptions I already said that.


Just so you know based on my example:
7.87 = 100%
5.99 = 76%
Delta = 24% aka XSX is 24% slower then PS5.

Please note that while xsx has nearly 6.0GB/s compressed data in this case the PS5 has only 7.9GB/s which is still not what Cerny said.
I mean this is simple math you can probably tell me if I've made any mistakes.

Correct what? you invent crazy PR-based math to deceive yourself.


My math:

1: PS5: 22GB / s
xbox one SX: 4.8GB / s

2: DF: "Resolution doesn't matter anymore because of asset geometry"

3: Asset geometry: SSD IO

So, PS5 with 5x more IO Band on the SSD will have 5x more geometry that equates to "5x more resolution"

PS4: 2160p

2160/5 = 432p

SX: 432p


Did you like my matemagic? Did you like my delirium? I copied yours.

 

[Bonsaibäumchen]

How do you get to 5.99/6.0 GB/s again, is this the made up 2-3x multiplier or taking the theoretical peak of the BCPack decompressor? If you do that I will take 22 GB/s for the Kraken one in the PS5 please .

First of - I already posted that if you'd cared to read.
I get there if i assume that BCPack can compress textures by 60%. ( this is what MS says they can do! )
100GB (Textures) *0.4 = 40GB
40GB / 2.4 GB ( SSD Speed ) = 16.66 seconds
So you've transferred 100GB worth of Data in 16.66 seconds = 100GB / 16.66 = ~6.00GB/S compressed data

btw. im a Sony Fan. Im just showing what that would mean.
It would mean that SXS is still slower by at at least 24%

And its a calculation in favor of xsx. So what? I don't care.
XSX Storage is still 24% slower then PS5
PS5 has ~14.x% less TFLOPS then XSX.

 

[Ascend]

How do you get to 5.99/6.0 GB/s again, is this the made up 2-3x multiplier or taking the theoretical peak of the BCPack decompressor? If you do that I will take 22 GB/s for the Kraken one in the PS5 please .

If you use 22GB/s that means that Kraken's compression would be 75%. We know it doesn't do that.
BCPack can probably reach 60%, but that's likely lossy compression rather than lossless. We'll see.

 

[thicc_girls_are_teh_best]

I don't know how I can make my post easier for you to understand. I watched NX's video several times. Why don't you take the time to READ my post.



I never said YOU did.

See, this is why you need READ my post. I made a list of accusations by fanboys who are attacking the PlayStation 5.




And?
I'm just over here shaking my head. You're missing the point. People are saying it's a 9.2TF console and not a 10.2TF. They were not referring to real world cases, they were talking about the actual TF count. This means if anyone says PS5 is a 10.2TF console, they would say, "No, it's really 9.2TF".


If you're not going to read my post, then dont waste my time, dude.

If you think people are trying to attack PS5, then ignore them. But I fail to see how their own actions are my responsibility, and mentioning them and what they do in relation to a response to me either insinuates you somehow think I'm responsible for that or enabling it in some way, or engaging in it myself. But you've already said the latter isn't what you're saying, and if you want to then say the former isn't either, then...why did you bring that up in the first place when I just said the systems are custom RDNA2 and therefore won't have nor need all the same RDNA2 features the PC GPUs will have?

The truth is I don't think your post is making that much sense; this seems to have all started because I made some responses to a post of yours that was talking about the Xbox community as a monolith engaging in FUD. But then you mentioned things to a way in which it's suggested anyone who speculates about things regarding PS5 and XSX to reach similar conclusions to the bullet points, somehow has an agenda to spread FUD against PS5.

And that's where I disagree; speculating that the SSD I/O delta between the two systems in real-world use-cases could be smaller than what the paper specs show isn't trying to hurt or damage the PS5; it's simply acknowledging the reality that there are multiple routes to solving the same challenges and resolutions involve hardware AND software implementations. Is speculation on PS5's GPU having higher clocks to help it punch above its weight with things affected by faster clocks (like pixel fillrate) somehow an attempt to hurt/damage the XSX and make it seem weaker? I don't think so, personally. So I would say the inverse, as long as it's not being done maliciously, is just healthy speculation and part of an optimistic outlook.

Saying that the systems won't reach the absolute stated theoretical computation limits isn't even in the same league as suggesting PS5 is "really a 9.2 TF system". We know the theoretical limit is higher than that, but we also know both systems won't reach their stated TF limits in most real-world situations. That's what I'm speaking on here, it shouldn't imply anything other than that.

If you want Xbox fans who ARE engaging in any malicious FUD spreading against PS5 to stop, then realistically I'd figure you'd also want PlayStation fans doing similar against XSX to stop. Because realistically that's the only way it COULD stop. However, the actual reality is that there will always be a few people on both sides who do this, it just comes with the territory. But as long as you yourself want to discuss things in a genuine way, then that's ultimately what matters. And FWIW, of what I've seen in this thread, I don't think I can say there are any great deal of Xbox fans or PlayStation fans engaging in malicious FUD attacks against their opposite systems. So I'm a bit lost as to why you started talking about people doing such in the first place.

 

[Deto]

If you use 22GB/s that means that Kraken's compression would be 75%. We know it doesn't do that.
BCPack can probably reach 60%, but that's likely lossy compression rather than lossless. We'll see.

compact yes in ideal cases, same as the ideal case of the SX to reach 6GB / s

but I guess:

MS always has the maximum theoretical peak in everything.

- 6GB / s
- 12TF

Sony NEVER has the theoretical maximum:

- 9TF
- 7GB / s SSD

 

[DForce]

If you think people are trying to attack PS5, then ignore them. But I fail to see how their own actions are my responsibility, and mentioning them and what they do in relation to a response to me either insinuates you somehow think I'm responsible for that or enabling it in some way, or engaging in it myself. But you've already said the latter isn't what you're saying, and if you want to then say the former isn't either, then...why did you bring that up in the first place when I just said the systems are custom RDNA2 and therefore won't have nor need all the same RDNA2 features the PC GPUs will have?

The truth is I don't think your post is making that much sense; this seems to have all started because I made some responses to a post of yours that was talking about the Xbox community as a monolith engaging in FUD. But then you mentioned things to a way in which it's suggested anyone who speculates about things regarding PS5 and XSX to reach similar conclusions to the bullet points, somehow has an agenda to spread FUD against PS5.

And that's where I disagree; speculating that the SSD I/O delta between the two systems in real-world use-cases could be smaller than what the paper specs show isn't trying to hurt or damage the PS5; it's simply acknowledging the reality that there are multiple routes to solving the same challenges and resolutions involve hardware AND software implementations. Is speculation on PS5's GPU having higher clocks to help it punch above its weight with things affected by faster clocks (like pixel fillrate) somehow an attempt to hurt/damage the XSX and make it seem weaker? I don't think so, personally. So I would say the inverse, as long as it's not being done maliciously, is just healthy speculation and part of an optimistic outlook.

Saying that the systems won't reach the absolute stated theoretical computation limits isn't even in the same league as suggesting PS5 is "really a 9.2 TF system". We know the theoretical limit is higher than that, but we also know both systems won't reach their stated TF limits in most real-world situations. That's what I'm speaking on here, it shouldn't imply anything other than that.

If you want Xbox fans who ARE engaging in any malicious FUD spreading against PS5 to stop, then realistically I'd figure you'd also want PlayStation fans doing similar against XSX to stop. Because realistically that's the only way it COULD stop. However, the actual reality is that there will always be a few people on both sides who do this, it just comes with the territory. But as long as you yourself want to discuss things in a genuine way, then that's ultimately what matters. And FWIW, of what I've seen in this thread, I don't think I can say there are any great deal of Xbox fans or PlayStation fans engaging in malicious FUD attacks against their opposite systems. So I'm a bit lost as to why you started talking about people doing such in the first place.

And you're still not able to understand the point of my post.

But I fail to see how their own actions are my responsibility, and mentioning them and what they do in relation to a response to me either insinuates you somehow think I'm responsible for that or enabling it in some way, or engaging in it myself.

Want to point me where I said it was your responsibility?


Right, I never said it.

Saying that the systems won't reach the absolute stated theoretical computation limits isn't even in the same league as suggesting PS5 is "really a 9.2 TF system".

It has nothing to do with my post.

I'm not going to repeat myself. Either you're trolling or your reading is just that bad. lol

Go over my post again before replying to me.

 

[Deto]

How do you get to 5.99/6.0 GB/s again, is this the made up 2-3x multiplier or taking the theoretical peak of the BCPack decompressor? If you do that I will take 22 GB/s for the Kraken one in the PS5 please .

The same as always with spoiled little children, who make up rules for them to win.

"in SX it's always the maximum theoretical peak"
"on PS5 you can never use the theoretical maximum"


Xbox: all my dreams and delusions are facts.
PS5: even facts are fake news.

 

[Trueblakjedi]

compact yes in ideal cases, same as the ideal case of the SX to reach 6GB / s

but I guess:

MS always has the maximum theoretical peak in everything.

- 6GB / s
- 12TF

Sony NEVER has the theoretical maximum:

- 9TF
- 7GB / s SSD

6GB/s is a decompression rate. Meaning if there are 6Gbs+ of data, it can decompress at that rate. A conversion rate.
12TF is a locked performance rate. clock * CU. There are never times when the frequency of the GPU will be less than 1.825 no matter what you throw at it.

The 9TF comes from the variable clock rate of the hardware... that's not the public's fault... thats Sony's design. Now its disingenuous to claim 9tf, but <10.28 tf isnt because there are times when the frequency of the GPU will not be 2.23 GHz depending upon what they throw at it.
7GB/s a system I/O which we know will vary... but not because of its decompression block capabilities which are truly breathtaking.

I'm not understanding the sturm and drang about these numbers here. Very emotional.

 

[Panajev2001a]

If you use 22GB/s that means that Kraken's compression would be 75%. We know it doesn't do that.
BCPack can probably reach 60%, but that's likely lossy compression rather than lossless. We'll see.

What? How do you come up with such maths?

The only figures we have now are (vendors provided):

XSX:
* uncompressed SSD I/O = 2.4 GB/s
* estimated effective compressed (taking into account BCPack and lzma) SSD I/O = 4.8 GB/s
* theoretical BCPack decompressor unit max throughput = 6 GB/s


PS5:
* uncompressed SSD I/O = 5.5 GB/s
* estimated effective compressed (taking into account Kraken) SSD I/O = 8-9 GB/s
* theoretical Kraken decompressor unit max throughput = 22 GB/s

Again, these are their marketing figures from both MS and Sony, period.

 

[Bonsaibäumchen]

compact yes in ideal cases, same as the ideal case of the SX to reach 6GB / s

but I guess:

MS always has the maximum theoretical peak in everything.

- 6GB / s
- 12TF

Sony NEVER has the theoretical maximum:

- 9TF
- 7GB / s SSD

You know whats funny?
Even though its this way, Xbox Fans can feel like they have the upper hand and wear the powercrown.

Any yet it's highly likley Sony is gonna win this generation again.

 

[Ascend]

The same as always with spoiled little children, who make up rules for them to win.

"in SX it's always the maximum theoretical peak"
"on PS5 you can never use the theoretical maximum"


Xbox: all my dreams and delusions are facts.
PS5: even facts are fake news.

I guess it's from both sides. Because any time someone brings up SFS as a talking point for the Xbox, people say that the same thing has been done since 2013 and that it will thus give zero advantage. Which is really absurd. Microsoft is not so stupid as to advertise 2013 tech as a new feature.

 

[geordiemp]

6GB/s is a decompression rate. Meaning if there are 6Gbs+ of data, it can decompress at that rate. A conversion rate.
12TF is a locked performance rate. clock * CU. There are never times when the frequency of the GPU will be less than 1.825 no matter what you throw at it.

The 9TF comes from the variable clock rate of the hardware... that's not the public's fault... thats Sony's design. Now its disingenuous to claim 9tf, but <10.28 tf isnt because there are times when the frequency of the GPU will not be 2.23 GHz depending upon what they throw at it.
7GB/s a system I/O which we know will vary... but not because of its decompression block capabilities which are truly breathtaking.

I'm not understanding the sturm and drang about these numbers here. Very emotional.

We are back on 9 TF FFS

Fanboys cant give the other console just 1 thing that its better at, just 1.

Not only has their platsic box got to be most powerful, all others have to fail.

Luckily for most normal people they are easy to spot

 

[Panajev2001a]

I guess it's from both sides. Because any time someone brings up SFS as a talking point for the Xbox, people say that the same thing has been done since 2013 and that it will thus give zero advantage. Which is really absurd. Microsoft is not so stupid as to advertise 2013 tech as a new feature.

Better and more efficient != necessarily a 2-3x multiplier
Take the description of SF and SFS up with MS’s own engineers if you want more bombastic data proclamations...

 

[Trueblakjedi]

What? How do you come up with such maths?

The only figures we have now are (vendors provided):

XSX:
* uncompressed SSD I/O = 2.4 GB/s
* estimated effective compressed (taking into account BCPack and lzma) SSD I/O = 4.8 GB/s
* theoretical BCPack decompressor unit max throughput = 6 GB/s


PS5:
* uncompressed SSD I/O = 5.5 GB/s
* estimated effective compressed (taking into account Kraken) SSD I/O = 8-9 GB/s
* theoretical Kraken decompressor unit max throughput = 22 GB/s

Again, these are their marketing figures from both MS and Sony, period.

I wouldnt call them marketing. But this post is 100% true generally speaking.

 

[Trueblakjedi]

We are back on 9 TF FFS

Are you timdog ?

You are pulling my leg right? You see that I'm responding to someones post yes?

If you aren't going to bother to read then don't write buddy.

 

[geordiemp]

You are pulling my leg right? You see that I'm responding to someones post yes?

If you aren't going to bother to read then don't write buddy.

Yeah it was just some fun, dont take it to heart, I was mocking the 9 TF posters...better worded now.

 

[rnlval]

No, I am not claiming that. Official specs state that data moves at 2.4 GB/s raw and 4.8 GB/s compressed. Those are the numbers I'm using.

Sure, Andrew Goossen is fired and not working at MS anymore.

What? How do you come up with such maths?

The only figures we have now are (vendors provided):

XSX:
* uncompressed SSD I/O = 2.4 GB/s
* estimated effective compressed (taking into account BCPack and lzma) SSD I/O = 4.8 GB/s
* theoretical BCPack decompressor unit max throughput = 6 GB/s


PS5:
* uncompressed SSD I/O = 5.5 GB/s
* estimated effective compressed (taking into account Kraken) SSD I/O = 8-9 GB/s
* theoretical Kraken decompressor unit max throughput = 22 GB/s

Again, these are their marketing figures from both MS and Sony, period.

*more than 6GB/s* claim used by Andrew Goossen

 

[Panajev2001a]

I wouldnt call them marketing. But this post is 100% true generally speaking.

So, it is either 6 GB/s vs 22 GB/s or 4.8 GB/s vs 8-9 GB/s...

 

[Ascend]

What? How do you come up with such maths?

The only figures we have now are (vendors provided):

XSX:
* uncompressed SSD I/O = 2.4 GB/s
* estimated effective compressed (taking into account BCPack and lzma) SSD I/O = 4.8 GB/s
* theoretical BCPack decompressor unit max throughput = 6 GB/s


PS5:
* uncompressed SSD I/O = 5.5 GB/s
* estimated effective compressed (taking into account Kraken) SSD I/O = 8-9 GB/s
* theoretical Kraken decompressor unit max throughput = 22 GB/s

Again, these are their marketing figures from both MS and Sony, period.

Compression works by making file sizes smaller. In reality, the XSX will never be able to exceed 2.4 GB/s of data throughput, and the PS5 will never be able to exceed 5.5 GB/s of data throughput. When it is compressed, that same 2.4 GB/s and 5.5 GB/s becomes to 'advertised' 4.8GB/s and 8-9GB/s.

The thing is, compression is not a constant. Some files will inevitably compress better than others, and the compression algorithm also influences how much the compression will be. Lossy compression is generally more effective than lossless, but within lossy vs lossless themselves there are differences in the results.

If you want to reach 6GB/s of effective output, that means that your 2.4GB/s needs to have 6GB/s worth of data. The compression ratio would be 6 / 2.4 = 2.5 to 1. Another way of looking at it is, that the file size has been reduced from 6 GB to 2.4GB, which means 2.4/6 = 40% of the original file size, meaning compression is 1 - 40% = 60% compression.

If you do the same, you arrive at 75% compression ratio for Kraken if it needs to output 22GB/s, although normally it would only be able to reach between 30%-40%, based on the given 8-9GB/s value.

Percentages might not be 100% accurate, but you get the point, I hope.

 

[Andodalf]

snip

It’s just a basic ad hominem attack. It’s used to try and discredit an argument by vilifying the one who makes it. I say this person is bad, so their argument is bad, so if you have the same argument you must be bad.

 

[Panajev2001a]

Sure, Andrew Goossen is fired and not working at MS anymore.


*more than 6GB/s* claim used by Andrew Goossen

For the max theoretical decompressor unit throughput?

More than 6 GB/s meaning 6.1 GB/s or 60 GB/s?

 

[Three]

I believe they did say it was RDNA2 "based".



It seems like you're just saying:

A) There are other ways to do this, therefore
B) This is the same as the other ways.

I don't think that is sound reasoning. The whole point of that paragraph I quoted is that the way you make your decisions has an impact on how much memory you save with your streaming, and Sampler Feedback can provide substantial improvements. Answer these questions if you a have a moment:

1) Is it impossible to you that under the hood of Sampler Feedback is more efficiently acquired and more accurate information than these other methods?

2) If this is so trivial, why is Sampler Feedback exclusive to DX12 ultimate? Why even bother putting this next to VRS and updates to DXR? Why all these articles/videos explaining the value of having feedback from texture sampling?

3) What do you make of the claim from Microsoft that games running on XBox One X were only making use of 1/2 or 1/3 of what was loaded in memory? Was this measure selectively taken (for marketing purposes) of games not using PRT techniques? Or do you just doubt that Sampler Feedback is sufficient to close that gap in any noticeable way like they claim?

1) I'm saying this is not some hardware feature that is secret sauce.
2) I'm saying that most of that efficiency comes from PRT already because what they are comparing it to is games without it.

As for the questions

1) I have no doubt that SFS provides an improvement but this is not something that isn't in other APIs other than DX under a different name/method. Slightly more or less efficient maybe even but 2x or 3x not a chance in hell. When the same guy "says having said that PRT has existed before" you can bet your bottom dollar that is whats being referred to in that multiplier

2) because it's a feature name of DX and they are advertising it, why do you think? It's the same as advertising the Move Engine on the xbox one. It makes it seem like it has things not possible elsewhere and becomes a selling point. The same thing happened with DX11 against Vulkan during the xbox one launch. I still remember everyone arguing about whether Vulkan had x, y, z.

3) yes i suspect that those games were loading things into memory that weren't needed all the time. In fact few games make use of it especially with the old HDDs because you can run into problems loading them into memory and the GPU stalling (framerate issues) . Just like the 'silly' example given for the x10 multiplier this is just a case of how did they do on data management to begin with.
RAGE made good use of it in its own way, Doom 4 did in some way, but a large majority of xbox one games don't. Doesn't mean it wasn't possible in some capacity though.

 

[rnlval]

Correct what? you invent crazy PR-based math to deceive yourself.


My math:

1: PS5: 22GB / s
xbox one SX: 4.8GB / s

2: DF: "Resolution doesn't matter anymore because of asset geometry"

3: Asset geometry: SSD IO

So, PS5 with 5x more IO Band on the SSD will have 5x more geometry that equates to "5x more resolution"

PS4: 2160p

2160/5 = 432p

SX: 432p


Did you like my matemagic? Did you like my delirium? I copied yours.

Epic saturated your PS5 with data and it gimped into mostly 1440p and 30 fps. PS5 has a high-end single SSD solution with non-high end compute processing hardware.

 

[Trueblakjedi]

Sure, Andrew Goossen is fired and not working at MS anymore.


*more than 6GB/s* claim used by Andrew Goossen

Hilarious i immediately went to linkedin to check hahahaha

https://www.linkedin.com/in/andrew-goossen-531a5b12/?msgConversationId=6666763706942152704&msgOverlay=true

 

[Panajev2001a]

Compression works by making file sizes smaller

Is this “patronising day”?! Really?

 

[Ascend]

Is this “patronising day”?! Really?

If that's the part you're going to reply to, I have nothing more to say to you.

 

[rnlval]

For the max theoretical decompressor unit throughput?

More than 6 GB/s meaning 6.1 GB/s or 60 GB/s?

Who knows... Not my problem.

 

[ToadMan]

There are never times when the frequency of the GPU will be less than 1.825 no matter what you throw at it

This is not accurate.

The gpu has a power cap and that cap is driven by the cooling capability of the system and the power supply.

If that power cap is exceeded the gpu will very quickly go to 0Mhz because it will shut down to protect itself from damage.

Now you could quite reasonably respond to this by saying “yes but the developers would never release such code” and you’d be 100% right.

What your numbers miss is that there are a set of developers who are using this equipment to achieve a goal. It’s not a mindless mechanical system.

Quoting numbers is only one part of what gets a game onto a console.

(Oh and your description of the 6gb/s decompression doesn’t make sense either but I could just be misunderstanding you)

 

[Panajev2001a]

Who knows... Not my problem.

lol yes, as long as it helps XSX seem as fast as if not faster than the PS5 your job is done ?

Seriously, what is the point of saying that? More importantly, what is the likelihood that more than 6 GB/s means anything more than 6.1 to, being generous, 6.5 GB/s... ?!

 

[Panajev2001a]

If that's the part you're going to reply to, I have nothing more to say to you.

I can reply to the rest and will, just found it a bit uncalled for, but if you start with that you are taking me for an idiot... so no loss for me in this case if this is your attitude.

 

[Deto]

You know whats funny?
Even though its this way, Xbox Fans can feel like they have the upper hand and wear the powercrown.

I've thought about it:


1: PS5: 22GB / s
xbox one SX: 4.8GB / s

2: DF: "Resolution doesn't matter anymore because of asset geometry"
Therefore, what the new "resolution" is the geometry of the asset.

3: Asset geometry: SSD IO

So, PS5 with 5x more IO Band on the SSD will have 5x more geometry that equates to "5x more resolution"

PS4: 2160p

2160/5 = 432p

SX: 432p

Anyone can invent such nonsense in any subject.

But it is not difficult to imagine who will believe these nonsense, it will be the same side that believed in 2013 in hidden GPU, that in 2020 believed in the comments on youtube of the "son" of the Sony employee saying that the PS5 overheated and caught fire and that sony was desperate.

 

[Bonsaibäumchen]

Correct what? you invent crazy PR-based math to deceive yourself.



1: PS5: 22GB / s
xbox one SX: 4.8GB / s

2: DF: "Resolution doesn't matter anymore because of asset geometry"

3: Asset geometry: SSD IO

So, PS5 with 5x more IO Band on the SSD will have 5x more geometry that equates to "5x more resolution"

PS4: 2160p

2160/5 = 432p

SX: 432p

Ah dude you just don't understand anything.
Thats fine.
Im Out.

 

[Panajev2001a]

Compression works by making file sizes smaller. In reality, the XSX will never be able to exceed 2.4 GB/s of data throughput, and the PS5 will never be able to exceed 5.5 GB/s of data throughput. When it is compressed, that same 2.4 GB/s and 5.5 GB/s becomes to 'advertised' 4.8GB/s and 8-9GB/s.

The thing is, compression is not a constant. Some files will inevitably compress better than others, and the compression algorithm also influences how much the compression will be. Lossy compression is generally more effective than lossless, but within lossy vs lossless themselves there are differences in the results.

If you want to reach 6GB/s of effective output, that means that your 2.4GB/s needs to have 6GB/s worth of data. The compression ratio would be 6 / 2.4 = 2.5 to 1. Another way of looking at it is, that the file size has been reduced from 6 GB to 2.4GB, which means 2.4/6 = 40% of the original file size, meaning compression is 1 - 40% = 60% compression.

If you do the same, you arrive at 75% compression ratio for Kraken if it needs to output 22GB/s, although normally it would only be able to reach between 30%-40%, based on the given 8-9GB/s value.

Percentages might not be 100% accurate, but you get the point, I hope.

I know what the raw I/O speed is and what the “compressed data transfer rate” is supposed to mean (you basically calculate average compression ratio for the mix of data you are going to be loading and providing a number that tells you what the effective bandwidth would be if you had to transfer the same amount of data uncompressed... you move 2.4 GB of compressed data in 1s and assume the data is uncompressed to 4.8 GB as it being stored into RAM that is as if you transferred all that data uncompressed over a 4.8 GB/s data pipe).

Take it with MS when estimated average compression rates decided to quote an equivalent bandwidth of 4.8 GB/s (which factors in the compression ratio... MS did the calculations you made in this thread, checked with technical and legal once more and published 4.8 GB/s while you are going for close to 6 GB/s).

You are essentially saying MS is happy with people comparing the two consoles and see almost 2x bandwidth disadvantage... and for what reason exactly?!

 

[Ascend]

I can reply to the rest and will, just found it a bit uncalled for, but if you start with that you are taking me for an idiot... so no loss for me in this case if this is your attitude.

I like to start my posts with a statement we can all agree on. It was not meant to be condescending. Sorry if you felt that way. I kind of get annoyed when I make an effort in posting something and someone replies on pretty much the least important part. Comes over as nitpicking rather than having a conversation.

For the max theoretical decompressor unit throughput?

More than 6 GB/s meaning 6.1 GB/s or 60 GB/s?

It's most likely somewhere between 6 and 7. Otherwise they would've advertised the higher number. That's why it's also likely closer to 6.4 than 6.8 for example, otherwise they could advertise with 'almost 7 GB/s".

 

[Deto]

You are essentially saying MS is happy with people comparing the two consoles and see almost 2x bandwidth disadvantage... and for what reason exactly?!

This is the same reasoning of those who believe in the hidden GPU of the xbox one.

They will invent that only at the end of 2021 the MS will be able to release the hidden power of the SSD because it has an NDA with AMD, because the SSD power of the SX is from RDNA3 and the MS cannot release the maximum power of the xbox SX until AMD to launch RDNA3 for PC.

It will be the same as 2013, a "misterxmedia 2" will appear shortly


"SX's SSD makes 6GB / s, but MS wants to surprise Sony and announced 4.8GB / s. When MS announces the price of the SX it announces that it has upgraded the SX SSD to 6GB / s "



I wonder if they believe that shit or are just lying to trick others.

 

[rntongo]

compact yes in ideal cases, same as the ideal case of the SX to reach 6GB / s

but I guess:

MS always has the maximum theoretical peak in everything.

- 6GB / s
- 12TF

Sony NEVER has the theoretical maximum:

- 9TF
- 7GB / s SSD

To be very honest. The only numbers we should be going by are 8-9GB/s for PS5 and 4.8GB/s for XSX. The 22GB/s by Cerny was a theoretical max, he said up to 22GB/s. How often it happens we'll have to wait and see. On the other hand the over 6GB/s for XSX that Andrew Goossen mentioned , he'll have to explain further as well.

 

[funking giblet]

In either case, I still don't see how the 2x -3x multiplier can be discarded.

Next-Gen PS5 & XSX |OT| Console tEch threaD

I tend to agree that they will be fundamentally similar. It’s just curious to see the difference in processing’s power dedicated to that though. XSX uses 1/10th of a core while PS5 has a dedicated block that’s equivalent to 1-2 cores. I guess Sony is anticipating handling magnitudes more I/O ops...

www.neogaf.com

When an engine already handles indexing the assets in a scene or precomputes the atlas, SRS will not improve it by much.

 

[Trueblakjedi]

I know what the raw I/O speed is and what the “compressed data transfer rate” is supposed to mean (you basically calculate average compression ratio for the mix of data you are going to be loading and providing a number that tells you what the effective bandwidth would be if you had to transfer the same amount of data uncompressed... you move 2.4 GB of compressed data in 1s and assume the data is uncompressed to 4.8 GB as it being stored into RAM that is as if you transferred all that data uncompressed over a 4.8 GB/s data pipe).

Take it with MS when estimated average compression rates decided to quote an equivalent bandwidth of 4.8 GB/s (which factors in the compression ratio... MS did the calculations you made in this thread, checked with technical and legal once more and published 4.8 GB/s while you are going for close to 6 GB/s).

You are essentially saying MS is happy with people comparing the two consoles and see almost 2x bandwidth disadvantage... and for what reason exactly?!

I think its only messageboards that are doing comparisons with alarm. I don't think the companies themselves feel anything but comfort with their solutions. MS seems supremely confident in fact.

I do have the sense that if there is a little more headroom for a frequency boost on the XSX GPU they will do it before launch a la what they did with XBO GPU.

 

[JareBear: Remastered]

I think its only messageboards that are doing comparisons with alarm. I don't think the companies themselves feel anything but comfort with their solutions. MS seems supremely confident in fact.

I do have the sense that if there is a little more headroom for a frequency boost on the XSX GPU they will do it before launch a la what they did with XBO GPU.

What would be a realistic expectation as far as GPU boost, an even 2.0Ghz?

 

[Panajev2001a]

I think its only messageboards that are doing comparisons with alarm. I don't think the companies themselves feel anything but comfort with their solutions. MS seems supremely confident in fact.

I do have the sense that if there is a little more headroom for a frequency boost on the XSX GPU they will do it before launch a la what they did with XBO GPU.

Maybe, it would not be bad for them to do so if it were possible. Maybe both will, they both seem confident to me.

 

[rntongo]

What would be a realistic expectation as far as GPU boost, an even 2,0Ghz?

They added 50 MHz on the the Xbox One. With the cooling they have the XSX adding 150MHz doesn’t sound too bad.

 

[Panajev2001a]

They added 50 MHz on the the Xbox One. With the cooling they have the XSX adding 150MHz doesn’t sound too bad.

Interesting, but I do not see them breaking that barrier so close to going live. Why would they need to push silicon further? Still, we shall see... not sure why you count Sony out since you have yet to see their cooling solution .

 

[TBiddy]

Yes, Sony will up the clock speeds. Obviously. 2.6 GHz at least.

 

[Andodalf]

What would be a realistic expectation as far as GPU boost, an even 2,0Ghz?

speculation is near impossible without seeing what AMD has for desktops as a reference

 

[rntongo]

Interesting, but I do not see them breaking that barrier so close to going live. Why would they need to push silicon further? Still, we shall see... not sure why you count Sony out since you have yet to see their cooling solution .

The Xbox One had a BIOS update after launch. No way the PS5 is increasing any clocks. They already need to use AMD shift to hit 2.23GHz on the GPU.

 

[Panajev2001a]

The Xbox One had a BIOS update after launch. No way the PS5 is increasing any clocks. They already need to use AMD shift to hit 2.23GHz on the GPU.

I recall a frequency bump of about 100 MHz (?) or so for the CPU and somewhat less than that for the GPU just before launch and then one for the Xbox One S edition, not post launch increases though... I must be forgetting though... do you have a link? Back then they had a 40% gap to try to close a bit more, what about this time? Why risk failing units?

Still, about PS5, always depends on the baseline and the top level clock which could both rise a a bit (it depends on what the power profile they keep constant is... that determines average baseline and max clock... they specifically said they cap the GPU at 2.23 GHz, it is not silicon’s max clock) we shall see I guess.
Maybe there is a multiplier somewhere...

 

[Panajev2001a]

Yes, Sony will up the clock speeds. Obviously. 2.6 GHz at least.

Yep, thanks for the confirmation. Exciting .