• Hey, guest user. Hope you're enjoying NeoGAF! Have you considered registering for an account? Come join us and add your take to the daily discourse.

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

Status
Not open for further replies.

IkarugaDE

Member
IMO it's a great achievement for Sony to keep the secrets for so long. It reminds me on apple at the early 2010's. Today, an iPhone is completely uncovered weeks before the keynote. In case of Samsung there even was a leak of a banner from the Samsung website they normally wanted to upload the same day after the offical reveal of the S20. Content: Buy a S20+ or s20 Ultra and get the Galaxy Buds+ for free. As this banner was leaked there wasn't even known there will be another Galaxy Buds.

(If it's not pure marketing from Apple and Samsung)
 
Last edited:
IMO it's a great achievement for Sony to keep the secrets for so long. It reminds me on apple at the early 2010's. Today, an iPhone is completely uncovered weeks before the keynote. In case of Samsung there even was a leak of a banner from the Samsung website they normally wanted to upload the same day after the offical reveal of the S20. Content: Buy a S20+ or s20 Ultra and get the Galaxy Buds+ for free. As this banner was leaked there wasn't even known there will be another Galaxy Buds.

(If it's not pure marketing from Apple and Samsung)
i mean theres more leaks from Sony than MS
 

Gavin Stevens

Formerly 'o'dium'
Not sure I buy the Anthem thing. Seems a bit of a reach to post that when by all accounts anthem has had a massive overhaul. It’s anything but a sinking ship, not when they have worked hard to turn it all around. If he had posted that when they released news of them updating it then sure, but now...? No.
 
  • Thoughtful
Reactions: Isa
Cerny talked about audio RT because it's new thing, technology in games, what's wrong with that? how it's misunderstanding? do you think he's gonna spil the beans about RT on his first interview? you might say ps5 lacks BC because cerny didn't talk about it or Ryan didn't mention it at CES.

and you still doubt it? :messenger_grinning_squinting: like what he's have to tell you to make you believe it?

wee don't even know if that leak is real and even it's real that don't disprove anything about ps5 just because there's RT in xsex leak.

yes it doesn't have RT because Ryan didn't mentions it but later he says that they will share more features about ps5 in coming months by then ps5 probably have RT by that time if he mentions it because Cerny mentioning is not enough for you,
it's hard to tell that you're trolling? or really have flawed logic?

key word: I see, next time send email to sony and ask that they will talk about what you want them to make a list or something.

here i can prove it:
nnui7v6ke9941.jpg



Sorry, are we substituing interviews and articles now for real world game results? Neither Microsoft nor Sony have yet proven the hardware accelerated ray tracing they've spoken of is worth a damn. So far they've just talked about it. One seems to be leaning more heavily into it than the other, kinda like the way Sony are leaning more heavily into their SSD performance. There can be no argument that Sony has projected the utmost confidence in their SSD solution, even making direct comparisons to the best of what was available on the PC (at the time the statement was made). Microsoft has been doing that regarding their GPU.

But back to those articles, If we are going by that then according to Sony prior to the release of the PS3, all of the PS3's big games were supposed to be native 1080p, because that's exactly the message they were sending with their 1080p full HD marketing designed to make 360 look inferior, and then it turned out that many of Sony's biggest PS3 releases were in the 720p range also, similar to the 360.

And whatever happened to dual 1080p 120fps on the PS3? Sony said that one, too.

So history does suggest we must verify what these companies are saying. Microsoft also said before Xbox one and PS4 release they weren't giving up a certain amount of power, what was it 30% to PS4? did they not? And it turns out that for all intents and purposes they were because the hardware differences were such that PS4 having that clear edge was inevitable. In certain cases and game releases, it appeared to be more.
 

darkinstinct

...lacks reading comprehension.
Damn, Geoff, don't do this. Smelling a PS5 delay into 2021 incoming. Would certainly fit with not going to E3, not revealing PS5 right now and Corona virus uncertainties.
 

SlimySnake

Flashless at the Golden Globes
What is "normal mode"? NVIDIA boost isn't the same as AMD boost.
If you look at official specs from nvidia, take the boost clock, add another >200mhz, and that's your sustained average clocks. For whatever reason, nvidia does the opposite of AMD and undersells their Tf vs AMD's over inflated number. NVIDIA shows 1710 boost clock and 10tf for 2080 when in reality it sits at >2000mhz and ~12Tf.


Yup. AMD advertises max theoretical power while NVIDIA low balls by quite a bit.
Its actually 1950 mhz on my system with rare jumps to 2000 mhz but yeah, always add another 200 mhz.

I think nvidia does it to make amd tflops look worse than they are. At 1.95 mhz, the rtx 2070 would be a 9.1 tflops card on par with the 9.6 tflops 5700xt but people look at the 7.3 tf number and think its not even a contest.

I wouldn't be surprised if thry use it to hide the true power draw of the card.
 

Ellery

Member
Yeah I think sinking ship is much more likely, but he instantly realized that it is kind of toxic so he deleted it.
 

devilNprada

Member
No surprise, 2.0 is outdated and doesn’t make sense to use this in new / current models.

Help me understand though I have a 2019 900 and a 2018 A8E what does it mean for them? I take it the 900 will get a firmware update the A8E will not? Or is this strictly 2020 models (which I don't think they make A8'sin 2020).
 
So who was the idiot that started the GCN rumor again? After Microsoft saying specifically that they are using RDNA?

Dunno; didn't even pop up on my radar until I watched the vid. But yeah, it was FUD, similar to the people trying to say PS5 didn't have hardware ray-tracing even tho Sony (specifically Cerny) outright said it did (and not understanding how the Oberon benchmarks and tests were actually structured).

Probably the same idiots who keep referring back to the debunked github leak.

Github leak was never debunked; Matt and Klee just said "disregard it", but failed to provide any context. Further GPU benchmarks still show the same chip with slightly tuned settings with successive steppings, which have if anything validated at least a chunk of the leak (i.e Oberon as a chip is still pertinent to PS5).

I keep trying to tell you that people can safely disregard parts of the leak as being potentially illogical, but it's not 0% relevant to PS5. Don't know why that's so difficult to understand.
 
Last edited:

bitbydeath

Gold Member
Github leak was never debunked; Matt and Klee just said "disregard it", but failed to provide any context. Further GPU benchmarks still show the same chip with slightly tuned settings with successive steppings, which have if anything validated at least a chunk of the leak (i.e Oberon as a chip is still pertinent to PS5).

I keep trying to tell you that people can safely disregard parts of the leak as being potentially illogical, but it's not 0% relevant to PS5. Don't know why that's so difficult to understand.

Yes it has. Everything about it has already been proven either incorrect (No RT) or extremely unlikely. (Dual-GPU setup).

MS using GCN has just as much chance.
 

FERN

Member
agree. well spent money. one more Spider Man and they going to rake the profit out of them. and still they have Ratchet and Clank, Resistence series.
They were already making 1st party games for PS, I guess maybe some companies were sniffing around thinking about possibly snatching the studio up. Sony wasn't taking any chances of losing them.
 

devilNprada

Member
They were already making 1st party games for PS, I guess maybe some companies were sniffing around thinking about possibly snatching the studio up. Sony wasn't taking any chances of losing them.

I noticed on their site they are doing a lot of VR games these days. Maybe that's it.
 

PaintTinJr

Member
(IIRC) Insomniac played a very big part in developing and sharing cutting edge programming techniques for making full use of the PS3 hardware in its first few years. Sony acquiring them begs the question:
“Is there too much secret sauce in the PS5 - again like the ps3 - that needs a technically brilliant dev team to demystify for the industry?”

If Sony went the crazy route and did dual APUs, this would certainly be a welcome addition to get people using the full hardware fast. More likely, Sony have bought them specifically to build a game that doesn't have historical rasterization in the design and is intended for RT from the ground up.
 
[
Yes it has. Everything about it has already been proven either incorrect (No RT) or extremely unlikely. (Dual-GPU setup).

MS using GCN has just as much chance.

Jeez, dude, you haven't even been paying attention. The no RT part was explained in RT not being enabled for regression tests (the RT blocks were disabled). The dual-GPU stuff is not even hinted in the Github leak; that is what other people have speculated to try and boost up the performance. But even in that case, again, disabled blocks. Oberon likely has a chunk of CUs (like the 48 CUs listed in a chip graphic drawn and posted a while ago) disabled likely due to bugs (a recent Oberon stepping fixed a bug related to memory bandwidth). PS4 GPU data also had a chunk of its CUs disabled for testing PS4 compatibility, no reason PS5 wouldn't be the same.

Plenty of people have spoken about both of these things in depth (including myself); you are trying really hard at this point to willfully ignore it when most of the people referencing Github leak aren't even 100% buying every single thing it mentioned (we don't need to, because successive GPU benchmarks have shown further Oberon progression).

So no, it hasn't been debunked whatsoever. Certain parts of it are unrealistic (retail GPU @ 2GHz for example), but other parts of it are still valid and Oberon itself is very much related to PS5.
 
Last edited:

bitbydeath

Gold Member
[


Jeez, dude, you haven't even been paying attention. The no RT part was explained in RT not being enabled for regression tests (the RT blocks were disabled). The dual-GPU stuff is not even hinted in the Github leak; that is what other people have speculated to try and boost up the performance. But even in that case, again, disabled blocks. Oberon likely has a chunk of CUs (like the 48 CUs listed in a chip graphic drawn and posted a while ago) disabled likely due to bugs (a recent Oberon stepping fixed a bug related to memory bandwidth). PS4 GPU data also had a chunk of its CUs disabled for testing PS4 compatibility, no reason PS5 wouldn't be the same.

Plenty of people have spoken about both of these things in depth (including myself); you are trying really hard at this point to willfully ignore it when most of the people referencing Github leak aren't even 100% buying every single thing it mentioned (we don't need to, because successive GPU benchmarks have shown further Oberon progression).

So no, it hasn't been debunked whatsoever. Certain parts of it are unrealistic (retail GPU @ 2GHz for example), but other parts of it are still valid and Oberon itself is very much related to PS5.

You can’t just not enable RT, it’s either their or it’s not and it wasn’t. We’re talking hardware here not software.

And yes the leak mentions two GPU’s, and it’s proven to not be an upgrade. Maybe they’re each the BC1 and BC2 leak.

There is no mention of total CU count.
It also has ties to Arden which was once thought to be Xbox related. And what was the excuse for that again? Oh that’s right, maybe an intern wrote it up.

Now don’t also forget these tests were run after Sony already had the hardware in their hands. IE. DEBUNKED!
 

PaintTinJr

Member
Certain parts of it are unrealistic (retail GPU @ 2GHz for example), but other parts of it are still valid and Oberon itself is very much related to PS5.

Has the 'retail GPU' part been leaked/confirmed? And is that consistent with what we know of the original PS4 APU heat/power draw and the PS3 RSX heat/power draw in the original PS3? The reason I ask is that I thought the RSX was based on a Nvidia Quadro card’s heat/power draw. Would Sony be buying at a level where they would be getting Radeon Pro level GPU silicon inside the APU? Or are they definitely getting 'retail'?
 
You can’t just not enable RT, it’s either their or it’s not and it wasn’t. We’re talking hardware here not software.

And yes the leak mentions two GPU’s, and it’s proven to not be an upgrade. Maybe they’re each the BC1 and BC2 leak.

There is no mention of total CU count.
It also has ties to Arden which was once thought to be Xbox related. And what was the excuse for that again? Oh that’s right, maybe an intern wrote it up.

Now don’t also forget these tests were run after Sony already had the hardware in their hands. IE. DEBUNKED!

36 CUs was listed as "full chip" Arden being in ps5 chip results was an error from the script used. It's real. Better to say it's all out of context, outdated or whatever but it isn't fake or disproven at all.
 

bitbydeath

Gold Member
36 CUs was listed as "full chip" Arden being in ps5 chip results was an error from the script used. It's real. Better to say it's all out of context, outdated or whatever but it isn't fake or disproven at all.

It is disproven as it doesn’t match what is known, I also wouldn’t rule out fake given the large amount of ‘errors’.
 
Has the 'retail GPU' part been leaked/confirmed? And is that consistent with what we know of the original PS4 APU heat/power draw and the PS3 RSX heat/power draw in the original PS3? The reason I ask is that I thought the RSX was based on a Nvidia Quadro card’s heat/power draw. Would Sony be buying at a level where they would be getting Radeon Pro level GPU silicon inside the APU? Or are they definitely getting 'retail'?

No no, I was just saying that even if Oberon (more likely a future stepping of Oberon) is the final chip, it very likely won't hit 2GHz in the retail model because that could balloon the power dissipation. And 2GHz is way beyond Navi's sweetspot for watts/performance, drawing tons more heat at that range for relatively poor (or at least tepid) gains in performance.
You can’t just not enable RT, it’s either their or it’s not and it wasn’t. We’re talking hardware here not software.

And yes the leak mentions two GPU’s, and it’s proven to not be an upgrade. Maybe they’re each the BC1 and BC2 leak.

There is no mention of total CU count.
It also has ties to Arden which was once thought to be Xbox related. And what was the excuse for that again? Oh that’s right, maybe an intern wrote it up.

Now don’t also forget these tests were run after Sony already had the hardware in their hands. IE. DEBUNKED!

If you've never done diagnostics testing on hardware, then you should at least know that certain parts of the hardware can be tested or not tested depending on what one wants to test. That's kind of what I was alluding to when mentioning the RT not being enabled; it simply wasn't be tested for in the regression tests.

...although, even with not providing that explanation, you are still wrong. If CUs can be enabled or disabled, then yes, RT cores can also be disabled or enabled. This should be plainly obvious.

No, Arden is still mainly thought to be XSX-related. There's a chance it could be related to Oberon but there haven't been any other GPU benchmarks datamined for Arden IIRC, just Oberon (out of the two). Of course no total CU count was mentioned, that's why I mentioned that there could be a chunk of CUs on Oberon that are not functional on prior steppings, could be related to bugs in the silicon. A semi-recent Oberon stepping fixed a silicon bug to up the bandwidth. And again, PS4 Pro's GPU during tests had a block of CUs disabled for PS4 regression testing.

What hardware? What are you even referring to anymore? I think you have a conspiracy that anyone who pays any interest to the Github leak either thinks PS5 will be 9.2TF @ 2Ghz or less than 9TF with the clock lowered. Well, talk to the people who reference to the leak in that capacity; I've said my piece plenty of times by now, I don't think either of scenarios are going to happen. Just dig through my post history on my personal speculations for most probable scenario regarding PS5 performance and what factors can lead to them. I'm not gonna type it all up again.
 

bitbydeath

Gold Member
No no, I was just saying that even if Oberon (more likely a future stepping of Oberon) is the final chip, it very likely won't hit 2GHz in the retail model because that could balloon the power dissipation. And 2GHz is way beyond Navi's sweetspot for watts/performance, drawing tons more heat at that range for relatively poor (or at least tepid) gains in performance.


If you've never done diagnostics testing on hardware, then you should at least know that certain parts of the hardware can be tested or not tested depending on what one wants to test. That's kind of what I was alluding to when mentioning the RT not being enabled; it simply wasn't be tested for in the regression tests.

...although, even with not providing that explanation, you are still wrong. If CUs can be enabled or disabled, then yes, RT cores can also be disabled or enabled. This should be plainly obvious.

No, Arden is still mainly thought to be XSX-related. There's a chance it could be related to Oberon but there haven't been any other GPU benchmarks datamined for Arden IIRC, just Oberon (out of the two). Of course no total CU count was mentioned, that's why I mentioned that there could be a chunk of CUs on Oberon that are not functional on prior steppings, could be related to bugs in the silicon. A semi-recent Oberon stepping fixed a silicon bug to up the bandwidth. And again, PS4 Pro's GPU during tests had a block of CUs disabled for PS4 regression testing.

What hardware? What are you even referring to anymore? I think you have a conspiracy that anyone who pays any interest to the Github leak either thinks PS5 will be 9.2TF @ 2Ghz or less than 9TF with the clock lowered. Well, talk to the people who reference to the leak in that capacity; I've said my piece plenty of times by now, I don't think either of scenarios are going to happen. Just dig through my post history on my personal speculations for most probable scenario regarding PS5 performance and what factors can lead to them. I'm not gonna type it all up again.

You’re bending far too much to try make it real, at some point the idea must snap.

Why would AMD even test already released hardware? They just get bored or did an intern accidentally hit the GO button?

Everything is against it from being related to a PlayStation product (read:Any)
 
Champions League returns on th 18th feb.
Sony is a huge sponsor.
Im betting some PS5 info in that week will drop.


this is true i remember watching the CL that year, they " see the future ads all over the advert screens around the pitch.. i could not find a picture of it on google. but they are huge sponsor, im expecting something this week from sony
 

LED Guy?

Banned
It wasn't just the E3 reveal video for scarlett. It's AMD's press release on their official blog where they commit to it while labeling it next gen RDNA. It's also the official Series X reveal post. Microsoft is doubling and tripling down. Cerny's first example was regarding audio ray tracing, no big deal, likely a misunderstanding. He then clarifies hardware acceleratio on a follow-up. I now think the issue is dead. Then the AMD GitHub leak where it doesn't appear at all for ps5 chip, but does for Xbox series x chip. Now I'm rethinking what Cerny said and now have doubts. Then comes CES, they put up a slide showing it. I again think the issue is put to bed, only to then bizarrely learn that of all features on the slide, the guy on stage literally talks about everything on the slide, except hardware based ray tracing.

Consider that to mention every other feature on that slide, he had to skip past hardware ray tracing twice. I found that extremely odd. Why do they seem so reluctant to talk about one of the most hyped new graphics features coming into next gen? Do I expect to see it in every game? No, it's too demanding. Now I've left open the possibility that they may have it, but I have legitimate reason to doubt it now. Phil, again, has further confirmed hardware accelerated ray tracing via interviews also.

So, long story short, I see one side projecting confidence whereas I see the other being evasive on the issue. Maybe I'm wrong, and I will gladly eat crow. But I'm going to need to see it in a released title to believe it.

And, make no mistake, just because I believe Xbox really has it, doesn't mean that they, too, don't need to prove it from where I'm sitting. Both need to prove they have the real deal, or they shouldn't claim that they do. That's how I see it. So hopefully that's a complete enough answer on my thought process. It'll be the last time I address it, though, till we get more verifiable information on both consoles, but Sony have not portrayed the same degree of confidence where ray tracing is concerned. If we are being honest, there has been a clear difference between how the two sides have been marketing the feature.
BROOOOOOOO!!!!

I do not know if you’re being real with yourself right now or not, Mark Cerny did talk about Ray Tracing as a visual effect and as a sound effect and this is according to the guy who wrote the WIRED article that it is associated with both of them, the sound and the visual, and Sony has CONFIRMED HARDWARE RAY TRACING TWICE FOR PS5, there’s no way around this my guy, give it up.

You’re starting to sound like one of those paranoid conspiracy theorists 😂😂
 
Last edited:
Status
Not open for further replies.
Top Bottom