Support NeoGAF

[ZywyPL]

The reports from right around the time Klee got confident about PS5 > XSX were that the XSX dev kits were coming late/having issues/etc.

36CU to achieve more than 12TF would have to run at ~2.7GHz, so that was false from the get-go. Maybe those PS>XB leaks actually refered to Lockhart instead?

 

[IntentionalPun]

36CU to achieve more than 12TF would have to run at ~2.7GHz, so that was false from the get-go. Maybe those PS>XB leaks actually refered to Lockhart instead?

I'm suggesting MS's dev kits might not have been 12TF, if dev kits weren't that powerful perhaps "insiders" went with that instead of leaks (up until the 12TF announcement from MS.)

 

[SlimySnake]

I dont think klee was lying. i dont think he was given false information.

i think sony changed the specs.

 

[Jon Neu]

But if Xbox flops more than PS5, is it a bad thing?

It's variable.

 

[EverydayBeast]

Some rumors have a shot of at least getting people talking about the consoles.

 

[thicc_girls_are_teh_best]

Well my theory has always been: the actual dev kits people were using. As others have pointed out we had "leaks" that XSX was 12TF from early 2019, but what about the devkits? The reports from right around the time Klee got confident about PS5 > XSX were that the XSX dev kits were coming late/having issues/etc.

Only problem with that part of Klee's thing is that if MS's devkits were having issues and coming in late, surely they would not have done the TGAs reveal when they did? And then posting the (early) APU shot on their Twitter account in early January. Now MS are basically saying the hardware is complete and they're just waiting for games to ship it out, while we haven't gotten the same said from Sony regarding PS5.

Though, they did drop a nice surprise with the controller today, so maybe they are "practically ready" to ship as well, readiness of launch games and any distribution issues (because of COVID-19) not permitting.

I dont think klee was lying. i dont think he was given false information.

i think sony changed the specs.

But from what is the question. If at some oint Sony considered going with a bigger chip but scaled back, when did that happen and how far into development on the bigger chip did they get before dropping it? Because it sounds like they decided on 36 CUs a long time ago, though they might've considered a 48 CU chip since Cerny used it as a hypothetical example.

If there was a larger PS5 chip that got into some type of silicon production for testing purposes, hopefully we find some proof of it in the future. For the record I don't think Klee was intentionally lying, either. Again, insiders can't control their sources, unless they are their own source. Maybe the numbers he got were from people using devkits with GCN-based cards in them? That is always a possibility.

 

[Genx3]

I'm suggesting MS's dev kits might not have been 12TF, if dev kits weren't that powerful perhaps "insiders" went with that instead of leaks (up until the 12TF announcement from MS.)

If this is true then why did Klee pretend to already know XSX was 12+TF's?
I remember after the Verge, Windows Central and DF announced that XSX was 12 RDNA2 TF's Klee chimed in with the he already knew that XSX was 12 Navi TF's.

 

[Genx3]

But from what is the question. If at some oint Sony considered going with a bigger chip but scaled back, when did that happen and how far into development on the bigger chip did they get before dropping it? Because it sounds like they decided on 36 CUs a long time ago, though they might've considered a 48 CU chip since Cerny used it as a hypothetical example.

If there was a larger PS5 chip that got into some type of silicon production for testing purposes, hopefully we find some proof of it in the future. For the record I don't think Klee was intentionally lying, either. Again, insiders can't control their sources, unless they are their own source. Maybe the numbers he got were from people using devkits with GCN-based cards in them? That is always a possibility.

The evidence points to PS5 always being a 36 CU SOC.
That is why the original rumor had PS5 listed at 8TF's.
Then GitHub leaked and PS5 was already up to 9.2 TF's.
Sony eventually tweaked the clocks all the way up to 2.23GHZ for 10.3 TF's
There was never a bigger SOC because Sony would have surely went with that one.

 

[Urprobablyright]

Any pro now... Would have a pro mk. 2 in a few years... So i don't think I'd get it.. maybe they should call it a PS5+ or something~ pay another $200 for a 2TB SSD or something~

I can't think of much they can upgrade really, like they seem to have decided that 32gb of ram is superfluous~

 

[Reallink]

Is it possible Sony had 2 SoC design candidates that were otherwise identical aside from CU counts? Think their 11th hour PS4 upgraded to 8GB. Perhaps unfavorable pricing trends in SSD's, GDDR, and Corona uncertainties forced them to downgrade to the cheaper part.

 

[Hawking Radiation]

We have Sony fanboys crying because PS5 isnt the grandaddy of TFlops and power. Even though they just decimated the competition this generation. Look at the output from Sony this gen. Its not good enough. They HAVE to be on top of everything. Then we have MS fanboys acting like because they have the power advantage, they have actually won something. Imagine being this down and out from an anemic ass generation that TFLOPs and gamepass is what gets you feeling like WINNER.

Never understood it.

It was NEVER about power. Last 30 years shows you its all about the software library. No one is crying over PS5 being lower power. Pretty much everyone is in agreement that Series X is the stronger console and because its the stronger console, that has to be reiterated time and time again throughout PS5 threads.

Start a thread about the innerworkings of the PS5's SSD and you'll come across a troll post barely 5 posts in. I'm speaking for myself when i say that i'm actually curious to know how this SSD works and what other design choices Sony made regarding the PS5 but people like myself seldom get the opportunity to discuss this with other more knowledgeable members because there are immature childish grown men who prefer to troll and spread FUD for the main aim of disrupting good conversation.

 

[rnlval]

Sorry. I do not understand your reply.

Can you explain to me? Thanks.

For XSX, MS claims an extra 13 TFLOPS equivalent from the RT cores. I recycled MS's claim for PS5 with adjustment for 36 CU and clock speed.

 

[LED Guy?]

I knew they were lying all along. Its obvious when people are being vague or just throwing up bs numbers.
People here are just gullible. They'll believe what ever they want to believe and someone came in and took advantage of that.
Back in 2013 when it leaked out that XB1 was 1.2 TF's (later upclocked to 1.3TF's) and PS4 was 1.8 TF's I knew that MS had messed up. The same way it leaked out that PS5's clocks were being tested at 2 GHZ with 9.2 TF's on tap (later boosted to 2.23GHZ and 10.3TF's). I knew that Sony had messed up and started pushing the clocks way up to try to close the gap.
It really is as simple as that.

I wouldn't say "Sony messed up", because 10.3 TF compared to 12 TF is only 16% difference, where as the differences between Xbox One & PS4 in GPU were almost 40% and XB1 had DDR3 memory, it was way worse.

Both Sony & Microsoft made very powerful consoles this time around and they've far exceeded our expectations, but I can say Microsoft went more with the GPU and Sony went extreme with the SSD speeds to truly make new worlds instead of pushing some more pixels, like the differences in real-world between them will be around 5 to 7 frames or 2160p for XSX and 2000p for PS5 which is almost nothing.

 

[Ashoca]

I wouldn't say "Sony messed up", because 10.3 TF compared to 12 TF is only 16% difference, where as the differences between Xbox One & PS4 in GPU were almost 40% and XB1 had DDR3 memory, it was way worse.

We don't know yet though, it's still 10.28 TFLOPs, 36 CUs at 2.23GHz (variable frequency), so it will go down. We don't even know the base clock, so the difference could be bigger.

Also, when you compare all numbers:

PS5 XBSX

Cores:
PS5: 2304
XSX: 3328

TMUs:
PS5: 144
XSX: 208

Rops
PS5: 64
XSX: 80

Bus:
PS5: 256
XSX: 320

That's a substantial difference between the two platforms for those trying to push that the only difference is 18% and limited to CU count.

So, XSX gives:
43% more cores.
44% more TMUs
25% more tops
25% more Mem Bus.

Also, you can still do a lot more work with 2TF's RDNA2 than you can with 500GF's of GCN, because nowadays the GPUs are way more efficient, i.e. a developer can do with 1 TFLOP of RDNA2.0 MUCH more than with 1 TFLOP GCN.

Anyways, we will see when the games arrive, how vastly different they will look like.

 

[LED Guy?]

I would not like this response to be interpreted as an attack on anyone in particular. It is not like this. I simply take advantage of the moment to try to contribute my point of view, especially in the general way that it is when it comes to judging information (or sometimes people).

Since I cannot speak for others, I will speak for myself. This was going to be a long text, but I've finally narrowed it down. Since I have been quoted, I will make a general reflection, although I think I already mentioned it once.

All truthful information is only true at the right time and under specific circumstances.

On the one hand there are the Specs of DK1, DK2, DKF and Retail. All different from each other. And people really want to know about retail. But since he cannot know them, he begins to speculate based on the DK. That is your responsibility. My responsibility is to say nothing that I cannot say. And assume the consequences of what I say.

That said, and as you all know, I only have the opportunity to discuss my personal experience with the DK1, which I consider to be top-tier information, but having the possibility does not mean that I am allowed to do so (NDA). That has always been clear. Beyond that, the rest of the information is second level, that is, through the mouth of my colleagues, who are subject to their own NDA. And then comes third-level information, contacts at other companies. Then the fourth level information would come, which would be that of the type "a type related to the sector (journalist, screenwriter, etc.) has contact with a worker from a developer and has told him that ...". And finally, in my opinion, would come the start of the dubious information, which would be the type "I have a butcher friend who has a cousin who works in ...". This is where the information has to be taken with the tips of your fingers because you can get burned. The rest of levels, 1st, 2nd, 3rd, up to 4th, is reliable information. Nobody in their right mind, and who wants a stable job, will risk their position for free (within those 4 levels) by giving information that can discredit them. It is possible that within the 4th group there is some erroneous information, but I do not think that deliberately.

Within the group of real Insiders (workers in the first group) there will be no one who comes to a public forum to say anything they cannot say. What we can do is talk about what has already been made public, as long as said conversation does not conflict with other unpublished information. And we can also comment on filtered information, as long as said conversation does not harm third parties. Once all these restrictions lose effect, over time, we can talk about everything. Which is what interests me.

It would also be helpful to make it clear that there are still things that have not been said to begin with. Therefore, branding any possible Insider a liar for the "Road to PS5 aimed at developers" would be a mistake, in my opinion. Surely there will have been one. And others don't. But come on, I think everyone has said what they have been able to say based on each of their specific circumstances. But I doubt that reputable and old users have put their reputation at risk for a time of "glory" knowing what was going to happen. It is clear to think that these people have either been deceived or were telling the truth at that "moment". Because the information goes from "moments".

We are going to give examples.

If you have a January 2019 Leak, it is legitimate to debate that Leak in 2020. But is it correct? Yes, but risky. You run the risk that this Leak, which was correct in 2019, is no longer correct in 2020, and therefore you are wasting time debating about it. Is it correct to debate the heating problems of a system in the DK1 of 2019 if it turns out that when you are debating it, that problem no longer exists or will it be corrected in the Retail version? Yes, it is correct to debate, but depending on how it can be an absurd waste of time using old information, as much as such old information was leaked yesterday. You decide what to debate and what not. And what to believe and what not. But it would not hurt to use common sense a little from time to time. Many times you fight to know that Devkit is the most powerful without stopping to think that this is of little importance with respect to Retail versions. And the specs of the Retail versions, which is what you really want to know, you will not know until their respective companies so wish.

But yes, this is a forum and you have to talk about something. But a top tier Insider will always tell you what he can talk about, and maybe what he can talk about is expired information. An insider here will be more useful once the consoles are released. Meanwhile, what you need is not an Insider. You need a spy or leaker.

As I have done so far, beyond the obvious jokes, it has been giving you indirect information that you could interpret correctly or not. For example, "puberty". What happens if you do the conversion from 9.2 RDNA2 to GCN? What number results?

Then there is my claim that PS5 was no longer 9.2. Given that I think there is no doubt either.

I know what you want. But what you want an Insider cannot give you right now. So I personally do not find it useful.

Health for everyone.

Thanks for this huge thorough write-up, now I understand everything about all this.

So maybe PS5 was at one point a 13 TFlops console, but that DK was an old one using GCN architectures and whatnot and some insider made a mistake about it.

 

[LED Guy?]

We don't know yet though, it's still 10.28 TFLOPs, 36 CUs at 2.23GHz (variable frequency), so it will go down. We don't even know the base clock, so the difference could be bigger.

Also, when you compare all numbers:

PS5 XBSX

Cores:
PS5: 2304
XSX: 3328

TMUs:
PS5: 144
XSX: 208

Rops
PS5: 64
XSX: 80

Bus:
PS5: 256
XSX: 320

That's a substantial difference between the two platforms for those trying to push that the only difference is 18% and limited to CU count.

So, XSX gives:
43% more cores.
44% more TMUs
25% more tops
25% more Mem Bus.

Also, you can still do a lot more work with 2TF's RDNA2 than you can with 500GF's of GCN, because nowadays the GPUs are way more efficient, i.e. a developer can do with 1 TFLOP of RDNA2.0 MUCH more than with 1 TFLOP GCN.

Anyways, we will see when the games arrive, how vastly different they will look like.

What you said is very wrong, we are comparing RDNA 2 to RDNA 2 so the comparison is as valid as GCN to GCN comparisons, it is exactly like that, your comparison works when PS5 is 10.3 GCN VS 12 RDNA 2 which is not, so the differences between them is 16% and the clocks are variable based on load, not based on temperature, so when you look at the sky, you would see the clocks are down to 500 Mhz for GPU because it isn't demanding place, but when you are at the most demanding place, the clocks will kick up to 2230 MHz instantly all the time, but the power consumption is stable.

The cores, ROPs, TMUs etc etc.... comparisons you've made are also not as you stated them to be, the PS5's GPU is clocked way higher than XSX, hence why it is a 10.3 TF compared to 12 TF, you are obviously trying to make it look huge when it really isn't.

This next-generation is the smallest differences we've ever seen in any generation.

 

[Bitmap Frogs]

OK, I just want to start off by saying: I know I will be getting many drive-by posts & many fanboys here arguing in this thread and insults thrown around, but please, anyone who does that will be reported as always, and @Mod of War has done a great work snuffing those trolls out of any threads.

Now that we have done with this, I just want an open & honest conversation about how the insiders and even journalists were wrong about the GPU of the PlayStation 5, individuals like @Kleegamefan @o'dium @OsirisBlack @HeisenbergFX4 and we can even throw the guy at WCCFTech who leaked the AMD Navi RDNA architecture way back in 2018 (he was right) and he said that PS5 & XSX are using Big Navi which exceeds 40 CUs, it seems like XSX is using it because it has 52 CUs while PS5 has 36 CUs for an overall of 12 TFlops and 10.3 TFlops respectively, so he was wrong with PS5 being Big Navi.

I mean there were many individuals who were strangely.....wrong, look at Kleegamefan, now I know everyone will say "Oh he's fake" and all that, but the guy's real name is Kelly Rickards, he used to work at GameFan and EGM back in the day, he has very string connections in the gaming industry, he even shared his name, E-mail address and all that with REEEE moderators like Daniel Ahmed (ZHuge) and others as well, but everything that guy said was true except the part where he said the PS5 is more powerful GPU-wise, he was right about both consoles being double digits and both being more feature packed and more powerful than AMD's best GFX card in 2019 which is RX 5700 XT, he also said the PS5's SSD is double the speed of XSX's SSD which turn out to be true before anyone knew or said.

There are many other examples, like:

1. Jason Schreier saying that both Sony & Microsoft were aiming for higher than Google's Stadia 10.7 TF number (no GCN/RDNA qualifiers).

2. Game Informer's Andrew Reiner, Former-IGN worker Colin Moriarty both were saying PS5 is more powerful.

3. We can even say BGs was stating that PS5 is in "puberty" and that it was 9.2 TF way back, 10 TFs isn't puberty, I know that for sure.

My speculation:

Maybe that Japanese journalist & VFXVeteran were right and Sony were going to make a PS5 Pro to launch along with the PS5 in 2020?

Hmm.....this does make you think a bit doesn't it? Maybe that's why insiders were wrong, they were leaking the PS5 Pro GPU specs?

I want your guys' opinions, please act civil about it, it's all fun talk here in the end, both consoles are very powerful and both far exceed our expectations, when someone spew those powerful specs that we have now in the consoles back in early 2019, everyone would laugh at that because they are very pricey and ridiculous, but look now.

That's my piece.

Let me break it down for you alright?

People who leak console warz info, be it real or made up are trolling for your attention. They are trolls. They might be professional trolls (journos) or amateur trolls (forum posters).

By making this thread you are giving them a second round of bonus attention after the specs have been revealed.

 

[BGs]

I think you should be less aware of the TFs. None will disappoint you. Although exceptional hardware is useless if the SDKs are not fully up to the task. Think more about what games you want to play and the choice will be easy. You are comparing a Ferrari and a Lamborghiny. You really are wasting your time. The console war as you knew it is coming to an end, it no longer makes sense, we are in an era where the most important thing will be software (games) over hardware. It is as if we came from the era of horse-drawn carriage. And let's move on to the automobile era. And people try to have conversations about horse-drawn carriages applying that law to the automobile. It is unsustainable, but you will realize it. You should be more aware of other details, such as the implication of the fact that the front light of the DS4 has disappeared in the DSense. That small detail is more important in the set than it can be to speak of idyllic TFs. Over time (maybe a lot) you will understand the "why".

On the other hand, I understand that from here to the consoles come out you have to talk about something, but really, I would take advantage of this time to open my mind to what is to come.

 

[Rightisright]

Some like Osiris, Schreier and Klee were just full of shit and tried to get a little bit of fame. Things and people have changed since klee was in the industry, and just because you were a writer for EGM doesnt mean you spoke to developers, you could just be writing reviews, or in charge of getting advertisements for the magazine.

Some however probably had early dev kit specs. Back when the early leaks were coming out, the initial dev kits were GCN GPUs (as RDNA 2 wasn't even able to be manufactured at that point), and were most likely higher in tflops.
Tommy Fisher had the XSX nailed, but the specs he put out for PS5 were for a GCN GPU due to how the L1 cache was structure was.
So he may have had old GCN numbers for PS5, but updated XSX specs.
But he was we'll off with his Remedy to PS and Techland to MS as well, so who knows.
But yes, most were just chasing the limelight, and NDAs are serious business.

 

[ZywyPL]

Also, when you compare all numbers:

PS5 XBSX

Cores:
PS5: 2304
XSX: 3328

TMUs:
PS5: 144
XSX: 208

Rops
PS5: 64
XSX: 80

Bus:
PS5: 256
XSX: 320

Source?

 

[Woo-Fu]

Aristotle postulated that nature abhors a vacuum. While that postulate wasn't very good in terms of physics it is great in terms of "news". There was a vacuum of ps5 news, people found creative ways to fill it.

 

[LED Guy?]

I think you should be less aware of the TFs. None will disappoint you. Although exceptional hardware is useless if the SDKs are not fully up to the task. Think more about what games you want to play and the choice will be easy. You are comparing a Ferrari and a Lamborghiny. You really are wasting your time. The console war as you knew it is coming to an end, it no longer makes sense, we are in an era where the most important thing will be software (games) over hardware. It is as if we came from the era of horse-drawn carriage. And let's move on to the automobile era. And people try to have conversations about horse-drawn carriages applying that law to the automobile. It is unsustainable, but you will realize it. You should be more aware of other details, such as the implication of the fact that the front light of the DS4 has disappeared in the DSense. That small detail is more important in the set than it can be to speak of idyllic TFs. Over time (maybe a lot) you will understand the "why".

On the other hand, I understand that from here to the consoles come out you have to talk about something, but really, I would take advantage of this time to open my mind to what is to come.

I agree with you a 100% there BGs, hope you have a good one!

But I gotta ask you a question, do you work for both Xbox Series X and PS5? If so, then can you tell us how good they are or maybe anything else about them?

 

[LED Guy?]

Source?

There isn't, he might link you to an Xbox fanboy saying it.

 

[ZywyPL]

There isn't, he might link you to an Xbox fanboy saying it.

Yeah, probably the one who also didn't answer:

Developers Begin to Weigh in on the Power Gap Between the Xbox Series X and PlayStation 5. Series X taking the power crown "from all accounts".

I'm a playstation guy but im intrigued about the differences in both architectures. At the end it comes down to price. All that supposed xbox power has to cost more unless microsoft wants to lose big money the first 3 years at least Do you think Sony's SSD costs the same as Microsoft's or...

www.neogaf.com

 

[Kumomeme]

The evidence points to PS5 always being a 36 CU SOC.
That is why the original rumor had PS5 listed at 8TF's.
Then GitHub leaked and PS5 was already up to 9.2 TF's.
Sony eventually tweaked the clocks all the way up to 2.23GHZ for 10.3 TF's
There was never a bigger SOC because Sony would have surely went with that one.

i dont remember where the rumors coming from..isnt it speculated based on cerny claim of 8tf minimum for running game at 4k? and there is view shared by Albert Penelo at another site where he personally think is 8tf

 

[BGs]

I agree with you a 100% there BGs, hope you have a good one!

But I gotta ask you a question, do you work for both Xbox Series X and PS5? If so, then can you tell us how good they are or maybe anything else about them?

In the video games section I only work for VR systems. It is all I can say about it.

Regarding how good is it? You mean games or systems? If you refer to hardware, I think I have already commented on it once, but if necessary, I repeat it. There is enough hardware (in both) to show photorealism. I think that should be enough to reassure anyone. I do not agree with those who comment that there will be no photorealism in the next generation (I know that recently a thread was created in this regard where the topic was discussed). If there is no photorealism in the next generation it will not be because the hardware is not capable of displaying it. Logically you have to be realistic, you cannot demand photorealism in 8k @ 60fps. But adjusting the resolution and frames appropriate for each experience, I bet there will be photorealism, with VR and without. Or at least the hardware capacity exists for it. Of course, it always depends on the skill, time and money of each company, but there is hardware capacity in both. It is also true that photorealism will be easier to achieve in some games than in others. Perhaps (for sure) the characters will be the most complex point to recreate, but seeing what ND's colleagues have achieved with TLOU2 I have no doubt that in the next generation there will be photorealistic characters in photorealistic environments (as much as to teach someone two photos and that does not know how to differentiate which is the real one). This does not mean that all the games are going to be photorealistic. Nor does it mean that it will be at the beginning of the generation. Logically it will not be so (the generation will be long). And perhaps during the whole generation there will be very few games like this, but it is that not everyone can afford a big budget and the best professionals. But there will be games that will be, without a doubt. The hardware allows it. Have I already said that the hardware allows it? (If any developer does not agree with me, we will talk about it again in a few years, if we are not all dead).

 

[Rightisright]

I'm was hesitant to brag, but it seems like I actually was the one with the insider status judging by results.
Did anyone get closer? All I had to do was say I had inside info and you would have all been praising me as legit.

"
PS5 -> 10.24tflops
XSX -> 12.2tflops"

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

I think the most important unknown right now is RAM both the type and the amount

www.neogaf.com

 

[Rightisright]

Well, I can argue that Matt was right when he said "glad that everyone had moved on from github" because the info shown was stating that PS5 is RDNA 1, 9.2 TF, no Hardware RT, no VRS, Navi 10 Lite, which the PS5 isn't, the PS5 is RDNA 2, Navi 2X, 10.3 TF, has HW RT and all that.

So I don't know what to put his words in the end.

As for Parrot AKA Thuway, well...he was right when Xbox One was having problems with Its OS and the fact that Microsoft lied when they revealed Xbox One's fake OS at their reveal event that took place on the 21st of May, 2013, and he even said Xbox One is a lot weaker in GPU than PlayStation 4 which is true, about 40% weaker, and he even stated that Xbox One will have ESRAM which it came true.

Kleegamefan everything he said was true minus the PS5 being more powerful in terms of GPU as he said "more flops".....he even strongly hinted at XSX being revealed at VGAs 2019 which became true, I wouldn't say he was BSing, he's a very nice guy and like I said, he shared so many things about him and his name and all that to ResetEra's mods.

All I just wanted to know is why everyone was so wrong? It seems......strange to me.

As everyone was saying about Github, it was real but lacked context. Context is, "what test were they doing?" "Would That test result need to mention RT or VRS?" "Is the 56CU XSX GPU showed the full chip that was shipping to devs (as devs use the full amount of CUs), or is it the retail chip that already has 4 Cu's disabled?"

That's all that was missing. And of course clock speeds arnt set until later either, so Sony could have decided after Github to push the clocks even further.

If you found yourself saying that Github was wrong, but fake insiders like Osiris or Klee were correct, then you have witnessed what confirmation bias is, and how easily you are prone to it.

 

[Genx3]

I wouldn't say "Sony messed up", because 10.3 TF compared to 12 TF is only 16% difference, where as the differences between Xbox One & PS4 in GPU were almost 40% and XB1 had DDR3 memory, it was way worse.

Both Sony & Microsoft made very powerful consoles this time around and they've far exceeded our expectations, but I can say Microsoft went more with the GPU and Sony went extreme with the SSD speeds to truly make new worlds instead of pushing some more pixels, like the differences in real-world between them will be around 5 to 7 frames or 2160p for XSX and 2000p for PS5 which is almost nothing.

I Agree that the PS5 wasn't nearly as bad as the XB1. The fact Sony went ahead and overclocked their console to close the gap is what I meant by mess up. MS did the same in 2013.
Percentages however don't tell the whole story without context.
So here is the context. A 40% GPU difference between PS4 and XB1 turned out to be 500 GCN GF's. The bandwidth difference between the PS4 and XB1 was some what mitigated by the 32MB's of ES Ram. Though PS4's GDDR5 solution was superior over all. The difference between PS5 and XSX is around 2 RDNA2 TF's. Could even be higher depending on the load of the PS5. The difference in Bandwidth for the GPU's in the XSX and PS5 is 112 GBPS. Again that is a significant difference.

2RDNA2 TF's and 112 GBPS Bandwidth are pretty significant.

i dont remember where the rumors coming from..isnt it speculated based on cerny claim of 8tf minimum for running game at 4k? and there is view shared by Albert Penelo at another site where he personally think is 8tf

The original Reddit rumors had a 4TF Lockhart, 8 TF PS5 and 12 TF Anaconda.
Then the fanboy madness ensued. All type of bs rumors that PS5 was 14 TF's, 13 TF's, etc.

 

[LED Guy?]

As everyone was saying about Github, it was real but lacked context. Context is, "what test were they doing?" "Would That test result need to mention RT or VRS?" "Is the 56CU XSX GPU showed the full chip that was shipping to devs (as devs use the full amount of CUs), or is it the retail chip that already has 4 Cu's disabled?"

That's all that was missing. And of course clock speeds arnt set until later either, so Sony could have decided after Github to push the clocks even further.

If you found yourself saying that Github was wrong, but fake insiders like Osiris or Klee were correct, then you have witnessed what confirmation bias is, and how easily you are prone to it.

I don't ever remember me saying that GitHub was wrong, I was just saying that it isn't the full context and that Klee and all those other guys were saying the full truth, but hey, no one was wrong nor right.

By the way, did you know that those GitHub folders containing next-gen consoles' specs were hacked/leaked off of AMD employee I think, and that guy is hmgqq who hosted those folders on GitHub, AMD didn't even put those folders on GitHub.

I can show you his tweets if you want, someone screen-capped it, I can show it to you if you want, his account now is protected by him, no one can see it.

 

[Genx3]

I'm was hesitant to brag, but it seems like I actually was the one with the insider status judging by results.
Did anyone get closer? All I had to do was say I had inside info and you would have all been praising me as legit.

"
PS5 -> 10.24tflops
XSX -> 12.2tflops"

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

I think the most important unknown right now is RAM both the type and the amount

www.neogaf.com

When your right your right!

 

[LED Guy?]

In the video games section I only work for VR systems. It is all I can say about it.

Regarding how good is it? You mean games or systems? If you refer to hardware, I think I have already commented on it once, but if necessary, I repeat it. There is enough hardware (in both) to show photorealism. I think that should be enough to reassure anyone. I do not agree with those who comment that there will be no photorealism in the next generation (I know that recently a thread was created in this regard where the topic was discussed). If there is no photorealism in the next generation it will not be because the hardware is not capable of displaying it. Logically you have to be realistic, you cannot demand photorealism in 8k @ 60fps. But adjusting the resolution and frames appropriate for each experience, I bet there will be photorealism, with VR and without. Or at least the hardware capacity exists for it. Of course, it always depends on the skill, time and money of each company, but there is hardware capacity in both. It is also true that photorealism will be easier to achieve in some games than in others. Perhaps (for sure) the characters will be the most complex point to recreate, but seeing what ND's colleagues have achieved with TLOU2 I have no doubt that in the next generation there will be photorealistic characters in photorealistic environments (as much as to teach someone two photos and that does not know how to differentiate which is the real one). This does not mean that all the games are going to be photorealistic. Nor does it mean that it will be at the beginning of the generation. Logically it will not be so (the generation will be long). And perhaps during the whole generation there will be very few games like this, but it is that not everyone can afford a big budget and the best professionals. But there will be games that will be, without a doubt. The hardware allows it. Have I already said that the hardware allows it? (If any developer does not agree with me, we will talk about it again in a few years, if we are not all dead).

Thanks again for answering my question, I'm so excited how technology in video games will be in next-generation and how PC technology will be progressing, from what you wrote in your comment, these consoles are looking very powerful, they remind me of Xbox 360 and PS3 being so powerful when they came out.

Thanks again and stay safe. ♥♥♥

 

[Genx3]

In the video games section I only work for VR systems. It is all I can say about it.

Regarding how good is it? You mean games or systems? If you refer to hardware, I think I have already commented on it once, but if necessary, I repeat it. There is enough hardware (in both) to show photorealism.

Games like Ryse and The Order of 1886 were already approaching realism on 1.3 TF and 1.8 GCN TF's consoles early this gen.

I could only imagine what could be done on 10TF and 12 RDNA2 TF consoles which are far more efficient and also include HW support for Ray Tracing.

Racing games, Flight games and other games with controlled environments should be able to hit close to photo realism without problems. Maybe even Open World games.

 

[LED Guy?]

I Agree that the PS5 wasn't nearly as bad as the XB1. The fact Sony went ahead and overclocked their console to close the gap is what I meant by mess up. MS did the same in 2013.
Percentages however don't tell the whole story without context.
So here is the context. A 40% GPU difference between PS4 and XB1 turned out to be 500 GCN GF's. The bandwidth difference between the PS4 and XB1 was some what mitigated by the 32MB's of ES Ram. Though PS4's GDDR5 solution was superior over all. The difference between PS5 and XSX is around 2 RDNA2 TF's. Could even be higher depending on the load of the PS5. The difference in Bandwidth for the GPU's in the XSX and PS5 is 112 GBPS. Again that is a significant difference.

2RDNA2 TF's and 112 GBPS Bandwidth are pretty significant.



The original Reddit rumors had a 4TF Lockhart, 8 TF PS5 and 12 TF Anaconda.
Then the fanboy madness ensued. All type of bs rumors that PS5 was 14 TF's, 13 TF's, etc.

No please, when we compare GPU/CPU/RAM differentials, we compare percentages, not how much numbers that are between them, if we were to go with your use case here, then the differences between PS4 and Xbox One are too big that Xbox One appears to be a last-gen console compared to it.

Compare an RTX 2080 against RX 5700 XT, the differences between them is AROUND 18%, and look at the performance differences, 7 to 9 frames in almost all games like that.

If you compare a 500 TFlops system and the other is 450 TFlops for example, the difference between them is a WHOPPING 50 TFLOPS when you 1st look at it, but it really is just a 10% difference.

About ESRAM, it was bottle-necked and barely reached Its full potential, this is why many games were hard to maintain against PS4 in terms of performance.

A 16% difference between PS5 & XSX is way smaller than the 40% difference that were between PS4 & XB1.

 

[Genx3]

No please, when we compare GPU/CPU/RAM differentials, we compare percentages, not how much numbers that are between them, if we were to go with your use case here, then the differences between PS4 and Xbox One are too big that Xbox One appears to be a last-gen console compared to it.

Compare an RTX 2080 against RX 5700 XT, the differences between them is AROUND 18%, and look at the performance differences, 7 to 9 frames in almost all games like that.

If you compare a 500 TFlops system and the other is 450 TFlops for example, the difference between them is a WHOPPING 50 TFLOPS when you 1st look at it, but it really is just a 10% difference.

About ESRAM, it was bottle-necked and barely reached Its full potential, this is why many games were hard to maintain against PS4 in terms of performance.

A 16% difference between PS5 & XSX is way smaller than the 40% difference that were between PS4 & XB1.

Percentages mean nothing without context.
2 TF's can factually do more computational work than 500 GF's. So the difference is significant.

 

[Rightisright]

When your right your right!

Just a guess, like all the other insiders lol.
MS had said the XSX was double One X, so that's 12, and add a touch more because I think MS would just be over the 12, so let's do 12.2.
Github showed a 36CU GPU, but I just could not see a way Sony would release a console under double digit tflops. They just couldnt do it. 9.2tflops would be suicide, so I assumed they might enable all 40CUs on the die at 2.0ghz, for 10.2tflops, and they would wear the lower yields.
Guess I'm a beer guesser than Osiris lol.

 

[LED Guy?]

Percentages mean nothing without context.
2 TF's can factually do more computational work than 500 GF's. So the difference is significant.

LMAO!! You are so misinformed I don't know where to even begin, the percentages are the most important figures, given they are the same GPU architectures (which they are).

16% difference will net you 16% more frames or resolution or more graphical prowess.

Whatever the XSX can do, PS5 can do it 16% lesser in resolution or whatever.

So, XSX at 2160p, PS5 at 2000p....those are the differences, I know you want the differences larger than they are but they aren't,

 

[prinz_valium]

Honestly? People lie. People lie all the time. Especially on the internet where you're anonymous.
Don't trust "leaks" that don't make sense and are vague.


Don't trust people who compare the Ps5 with the Series X instead of Ps5 with Ps4.
Don't trust people who don't have a proven track record and leaked correct and verifiable before.

No one on the Ps5 leaked anything right expect for the pictures of the dev kit. Those were the only real leaks.
Oh and GitHub of course, but "Insiders" claimed that's not correct for Ps5.
On the Series X side, everything turned out to be correct and came from trusted sources who have leaked first hand information in the past.



It's super easy to know how are fake insiders and who are real insider. People just like to believe bullshit.

 

[thicc_girls_are_teh_best]

Also, you can still do a lot more work with 2TF's RDNA2 than you can with 500GF's of GCN, because nowadays the GPUs are way more efficient, i.e. a developer can do with 1 TFLOP of RDNA2.0 MUCH more than with 1 TFLOP GCN.

Very important part to emphasize here; when you look at the GPGPU offloading programming Sony devs (and some 3rd parties) were able to do with 300 GF - 500 GF of headroom on the GPU (while still delivering visual parity with top-looking XBO games, if not moreso), it's quite staggering. A lot of the reason that trend came about is because that gen of consoles had dogshit CPUs to begin with (tho maybe that's being a bit harsh; they were marginally better than Cell in most areas (worst in some others), but it's not like CPUs haven't been used a ton in the past. The Motorola 68K is probably the most popular CPU of all time in gaming going by the number of systems that used it (MegaDrive, Neo-Geo, Atari Jaguar, Sharp X68000, etc.). But that CPU was revolutionary and ahead of its time in a lot of ways, such as having 32-bit registers. Hence why it was such a popular CPU).

Now you're dealing with an XSX with a 1.87 TF - 2 TF headroom over the PS5, on RDNA2 architecture. Plus ML, VRS, DLSS-based technologies baked into the silicon on top of that. That is a pretty substantive advantage for GPGPU programming tasks while still being able to provide visual and framerate parity with PS5 titles (at the very least). And to add to that, both consoles this gen actually have very respectable CPUs, so there's not nearly as much a need to offload CPU tasks to the GPU for sake of poor CPU performance, freeing up GPGPU programming potential for even more lavish and complex physics, AI, logic, etc. code.

As interested as I am in the SSDs, tbqh it's the GPGPU programming potential that's the real draw for paradigm shifts in next-gen game design IMO. And I think that brings about interesting advantages between the two systems: PS5 with its SSD (advantage in texture and data streaming, game load times etc.), XSX with its larger GPU (advantage in RT, visual throughput, GPGPU programming of parallelized AI/physics/logic etc. tasks). And that doesn't mean the other system is a slouch in those departments, either; BUT, they would have to make more compromises to match the given advantages of the other in those specified areas.

It will be very fun to see both systems pushed to their limit in a few years from now; even the launch titles should offer some tasty previews of their capabilities.

No please, when we compare GPU/CPU/RAM differentials, we compare percentages, not how much numbers that are between them, if we were to go with your use case here, then the differences between PS4 and Xbox One are too big that Xbox One appears to be a last-gen console compared to it.

Compare an RTX 2080 against RX 5700 XT, the differences between them is AROUND 18%, and look at the performance differences, 7 to 9 frames in almost all games like that.

If you compare a 500 TFlops system and the other is 450 TFlops for example, the difference between them is a WHOPPING 50 TFLOPS when you 1st look at it, but it really is just a 10% difference.

About ESRAM, it was bottle-necked and barely reached Its full potential, this is why many games were hard to maintain against PS4 in terms of performance.

A 16% difference between PS5 & XSX is way smaller than the 40% difference that were between PS4 & XB1.

No, percentages aren't the only metric to use when comparing systems, because different workloads require different amounts of computational power that can be measured in aspects besides percentages. So using percentages alone is not the best way to view system capabilities.

For example, the PS5's SSD's raw sequential read speed is rated at 5.5 GB/s, while XSX's is rated at 2.4 GB/s. That's about a 125% percentage advantage for PS5 there. However, this is sequential read speed, so the most obvious way to look at this as a practical example is in loading game data. If a given game can load its contents from the SSD into system RAM on XSX in roughly 1 second, then logically PS5 would load it in half a second.

But to the end-user, the difference in that speed as they perceive it will be barely noticeable. So you have a 125% advantage in sequential read speeds for the SSD but in practice you shave off half a second of loading for the next section of the game. All of the sudden that 125% delta doesn't look so impressive. That is just one use-case, however, and yes I'm simplifying things. But I just wanted to illustrate how you can't simply rely on highly general percentage differences alone (general as in, not factoring in percentage differences of other elements of the system that at the sub-level).

When you look at the percentages differences of sub-elements within a given system, you can better figure out what particular tasks the components those sub-elements comprise will benefit and by how much. Ironically, if you simply take the TF differential percentage as your main takeaway, you are indirectly saying TFs are the only thing that matters. Especially if you ignore the non-graphics orientated tasks GPUs can fulfill these days by way of GPGPU programming and other things.

 

[LED Guy?]

Very important part to emphasize here; when you look at the GPGPU offloading programming Sony devs (and some 3rd parties) were able to do with 300 GF - 500 GF of headroom on the GPU (while still delivering visual parity with top-looking XBO games, if not moreso), it's quite staggering. A lot of the reason that trend came about is because that gen of consoles had dogshit CPUs to begin with (tho maybe that's being a bit harsh; they were marginally better than Cell in most areas (worst in some others), but it's not like CPUs haven't been used a ton in the past. The Motorola 68K is probably the most popular CPU of all time in gaming going by the number of systems that used it (MegaDrive, Neo-Geo, Atari Jaguar, Sharp X68000, etc.). But that CPU was revolutionary and ahead of its time in a lot of ways, such as having 32-bit registers. Hence why it was such a popular CPU).

Now you're dealing with an XSX with a 1.87 TF - 2 TF headroom over the PS5, on RDNA2 architecture. Plus ML, VRS, DLSS-based technologies baked into the silicon on top of that. That is a pretty substantive advantage for GPGPU programming tasks while still being able to provide visual and framerate parity with PS5 titles (at the very least). And to add to that, both consoles this gen actually have very respectable CPUs, so there's not nearly as much a need to offload CPU tasks to the GPU for sake of poor CPU performance, freeing up GPGPU programming potential for even more lavish and complex physics, AI, logic, etc. code.

As interested as I am in the SSDs, tbqh it's the GPGPU programming potential that's the real draw for paradigm shifts in next-gen game design IMO. And I think that brings about interesting advantages between the two systems: PS5 with its SSD (advantage in texture and data streaming, game load times etc.), XSX with its larger GPU (advantage in RT, visual throughput, GPGPU programming of parallelized AI/physics/logic etc. tasks). And that doesn't mean the other system is a slouch in those departments, either; BUT, they would have to make more compromises to match the given advantages of the other in those specified areas.

It will be very fun to see both systems pushed to their limit in a few years from now; even the launch titles should offer some tasty previews of their capabilities.



No, percentages aren't the only metric to use when comparing systems, because different workloads require different amounts of computational power that can be measured in aspects besides percentages. So using percentages alone is not the best way to view system capabilities.

For example, the PS5's SSD's raw sequential read speed is rated at 5.5 GB/s, while XSX's is rated at 2.4 GB/s. That's about a 125% percentage advantage for PS5 there. However, this is sequential read speed, so the most obvious way to look at this as a practical example is in loading game data. If a given game can load its contents from the SSD into system RAM on XSX in roughly 1 second, then logically PS5 would load it in half a second.

But to the end-user, the difference in that speed as they perceive it will be barely noticeable. So you have a 125% advantage in sequential read speeds for the SSD but in practice you shave off half a second of loading for the next section of the game. All of the sudden that 125% delta doesn't look so impressive. That is just one use-case, however, and yes I'm simplifying things. But I just wanted to illustrate how you can't simply rely on highly general percentage differences alone (general as in, not factoring in percentage differences of other elements of the system that at the sub-level).

When you look at the percentages differences of sub-elements within a given system, you can better figure out what particular tasks the components those sub-elements comprise will benefit and by how much. Ironically, if you simply take the TF differential percentage as your main takeaway, you are indirectly saying TFs are the only thing that matters. Especially if you ignore the non-graphics orientated tasks GPUs can fulfill these days by way of GPGPU programming and other things.

I know it doesn't work on everything, but we've been using it to measure percentages between GPU that have the same architectures and assume that the rest of the console (or a PC in another example) to be the same, all I'm saying is that this is the most convenient way of comparing 2 GPUs against each other assuming the stuff I said before.

 

[S0ULZB0URNE]

For XSX, MS claims an extra 13 TFLOPS equivalent from the RT cores. I recycled MS's claim for PS5 with adjustment for 36 CU and clock speed.

In FP16 mode ie not the full FP32

 

[psorcerer]

OK, I just want to start off by saying: I know I will be getting many drive-by posts & many fanboys here arguing in this thread and insults thrown around, but please, anyone who does that will be reported as always, and @Mod of War has done a great work snuffing those trolls out of any threads.

Now that we have done with this, I just want an open & honest conversation about how the insiders and even journalists were wrong about the GPU of the PlayStation 5, individuals like @Kleegamefan @o'dium @OsirisBlack @HeisenbergFX4 and we can even throw the guy at WCCFTech who leaked the AMD Navi RDNA architecture way back in 2018 (he was right) and he said that PS5 & XSX are using Big Navi which exceeds 40 CUs, it seems like XSX is using it because it has 52 CUs while PS5 has 36 CUs for an overall of 12 TFlops and 10.3 TFlops respectively, so he was wrong with PS5 being Big Navi.

I mean there were many individuals who were strangely.....wrong, look at Kleegamefan, now I know everyone will say "Oh he's fake" and all that, but the guy's real name is Kelly Rickards, he used to work at GameFan and EGM back in the day, he has very string connections in the gaming industry, he even shared his name, E-mail address and all that with REEEE moderators like Daniel Ahmed (ZHuge) and others as well, but everything that guy said was true except the part where he said the PS5 is more powerful GPU-wise, he was right about both consoles being double digits and both being more feature packed and more powerful than AMD's best GFX card in 2019 which is RX 5700 XT, he also said the PS5's SSD is double the speed of XSX's SSD which turn out to be true before anyone knew or said.

There are many other examples, like:

1. Jason Schreier saying that both Sony & Microsoft were aiming for higher than Google's Stadia 10.7 TF number (no GCN/RDNA qualifiers).

2. Game Informer's Andrew Reiner, Former-IGN worker Colin Moriarty both were saying PS5 is more powerful.

3. We can even say BGs was stating that PS5 is in "puberty" and that it was 9.2 TF way back, 10 TFs isn't puberty, I know that for sure.

My speculation:

Maybe that Japanese journalist & VFXVeteran were right and Sony were going to make a PS5 Pro to launch along with the PS5 in 2020?

Hmm.....this does make you think a bit doesn't it? Maybe that's why insiders were wrong, they were leaking the PS5 Pro GPU specs?

I want your guys' opinions, please act civil about it, it's all fun talk here in the end, both consoles are very powerful and both far exceed our expectations, when someone spew those powerful specs that we have now in the consoles back in early 2019, everyone would laugh at that because they are very pricey and ridiculous, but look now.

That's my piece.

I thin it's pretty obvious that there were mutiple devkits with various specs.
Sony is claiming that dev support was a number one priority, so they probably "play-tested" a lot of variants.
A lot of custom silicon was stuffed in the PS5, there is no way to make it work for developers without a solid feedback.

 

[LED Guy?]

I thin it's pretty obvious that there were mutiple devkits with various specs.
Sony is claiming that dev support was a number one priority, so they probably "play-tested" a lot of variants.
A lot of custom silicon was stuffed in the PS5, there is no way to make it work for developers without a solid feedback.

That is exactly what I think happened, I think the PS5 had many dev kits ranging from 13 TFlops to 8 TFlopsm that is why we are getting mixed numbers all the time including Kleegamefan's leak as well.

 

[VFXVeteran]

It was obvious he was full of shit. Even his claim to fame about the PS4 Pro is quite clearly BS when you look into it.

The issue here is that what he said about a PS5 Pro and base, and their respective TFs was utterly wrong. He said that the base would be 9TF and Pro somewhere towards 12. Yet here we are with a PS5 confirmed pretty much bang between those numbers meaning that a budget version at - 1.2TF or a Pro at + 1.7TF are just not realistic. Sony have actually announced the perfect number of TFs to debunk either of those figures because a whole new system above or below to meet those numbers would be pointless.

VFXVeteran was right about Horizon Zero Dawn, but he's been very much wrong in regards to the PS5.

You are reading the last leak that I got verification about 2 SKUs on. My sources did indeed tell me that the PS5 base was 1080-1080Ti performance (before that post) and they was right amid all of the other "insiders" and console warriors saying the opposite.

The latest leak was my sources verifying to me that their would be 2 SKUs but didn't know when the other SKU would come out. Also, they never mentioned the TFLOPs of the other SKU. I guessed it was the PS5 Pro. It could have been that the other SKU is a lower end model like Lockhart.

I'm not wrong until Sony says they are only making one SKU. The sources that I have also told me about the PS exclusives coming to PC too. And I believe that was also correct (and will continue to be correct as there are more than just 1 PS exclusive coming to PC).

In any case, it doesn't matter as whatever I find out in the future I'll keep to myself. After all the abuse that good people have had to weather (i.e. DF, and other people working in the industry), I don't think anyone is ever going to feed the rude PS warriors with any inside info going forward.

 

[Genx3]

LMAO!! You are so misinformed I don't know where to even begin, the percentages are the most important figures, given they are the same GPU architectures (which they are).

16% difference will net you 16% more frames or resolution or more graphical prowess.

Whatever the XSX can do, PS5 can do it 16% lesser in resolution or whatever.

So, XSX at 2160p, PS5 at 2000p....those are the differences, I know you want the differences larger than they are but they aren't,

What I stated are facts.
Its not my fault they hurt your feelings.
By the way I never stated that the PS5 couldn't do everything the XSX can do at lower settings. Of course it can.

 

[LED Guy?]

What I stated are facts.
Its not my fault they hurt your feelings.
By the way I never stated that the PS5 couldn't do everything the XSX can do at lower settings. Of course it can.

Yes, it is just like that, the differences are so small and we should be happy about that.

 

[Bernkastel]

I will just leave this piece of art here.

the lead graphics engine designer laughs at me for stopping every 30 seconds to just rotate the camera and look at the sights and listen to the sounds of their game world. "Everything in the game is PBR with multiple sources of volumetric lighting. All of the characters in this game are unique. You will never see the same character model more than once." As I am playing I realize how difficult this game is. I play several games competitively multiple genres (Smite,R6,SFV,CoD,BF,League) but I am getting my ass handed to me. I take the lead character out into a hallway and two enemies rush into the room shooting and moving in what appears to be a tactical pattern. "Is this fucking guy shooting at me while the other guy is trying to circle around to flank me?" he laughs again. "Yep, suppressive fire." As the enemy shoots at me I take note of the deformation of the geometry on display, chunks of cement give way revealing the metal pipes underneath that hold the structure together. Everything is based in reality including what happens when you shoot people or get shot. Its beautiful but it aint pretty.

I sit the controller down (I'm about to die again anyways) "The game is definitely running at 60fps" he shakes his head at me "No its running at 90fps locked. native 4k." dumb stare "On which machine?" he smiles "Both of them."

Two hours and several drinks later

When I asked about the performance of both consoles his response was "Both have the same target we will see if either of them hit that target. " I pushed further and asked what the target was and his response was simply."Above 12" That is not for the XseX or the PS5,from what I can gather internally both machines are targeting very high performance which lines up with hearing earlier that the machines would both be $499 with both companies Taking a loss.

Notes
The gdk they are working with is not the final kit for either machine and XseX is currently ahead in terms of sheer power. Sonys tool set is much more complete and refined.
The trailer is all actual game play. Not in engine but game play.
Every developer or engineer I have ever met day drinks.

 

[BGs]

What I stated are facts.
Its not my fault they hurt your feelings.
By the way I never stated that the PS5 couldn't do everything the XSX can do at lower settings. Of course it can.

In practice you will not be able to say that one goes in Ultra and the other in High. Both will go in Ultra. If not, it would be a suspicious case worth investigating. Another topic will be resolutions and Fps. But the difference should be so ridiculous that it is absurd to worry about it. But I'm sure there will be cross-platform games that will benefit one system more than another. That has always existed and will always exist, but this time it will be more evident that it is not due to lack of power. To really appreciate the difference, you should consider the exclusive FirstParty.

 

[Dante83]

Whomever who leaked 13 TF of compute power is looking real silly right now. No one is going to take that dude seriously.

 

[BluRayHiDef]

I really don't want a PS5 Pro. Perhaps in 2023/4, but not this year. The PS4 Pro has a butterfly GPU resulting in double TFLOPs. A PS5 Pro should be at least double PS5... 20.6TFLOPs?

I don't see it and don't want it. SSD would feel too small, RAM and its bus would feel too small... The PS5 as it exists is too powerful for there to be a Pro launching along-side it.

If they release two PS5 SKUs at launch, the less powerful one should be designated simply PlayStation 5 and the more powerful one Playstation 5 Pro. Then, at a later point in the generation Sony can release another SKU designated PlayStation 5 Advanced Pro ("Ad Pro" for short).