Support NeoGAF

[SportsFan581]

Well you can also say that if the PS5 is 400, you might as well pay an extra 100 and get the better Xbox. It all depends on how people see things doesn’t it.

I think another reason for Lockhar’s existence is xcloud. They can use that in the servers because they are primarily targeting phones and tablets where 4K is largely pointless. While they are at it they can do a cheap retail version at the same time to cover more of the market.

I’m just guessing for fun. But one thing for certain is that those Codenames and chip tests are heavily tied to both PlayStation and Xbox hardware. That is undeniable.

Of course, It's just the TF numbers between Lockhart and PS will be drastically different. But, than Lockhart could retail for less than we think (that 4TF chip would be a very small APU compared to the chips in the higher-end models).

 

[Dee Dah Dave]

Of course, It's just the TF numbers between Lockhart and PS will be drastically different. But, than Lockhart could retail for less than we think (that 4TF chip would be a very small APU compared to the chips in the higher-end models).

It will also be 1080p whereas the PS5 will be ~4K, barring resolution the graphics will be largely similar. And there are still a huge amount of gamers with standard HD TV’s. So it’s not a completely silly idea and will probably hold it’s own quite well. It wouldn’t interest me but it will some.

 

[SportsFan581]

It will also be 1080p whereas the PS5 will be ~4K, barring resolution the graphics will be largely similar. And there are still a huge amount of gamers with standard HD TV’s. So it’s not a completely silly idea and will probably hold it’s own quite well. It wouldn’t interest me but it will some.

I can see it selling well, I've bought the last Xbox consoles straight out of the gate, so $500 is a typical price for me. But a lot of buyers wait on next gen until prices decrease (those first good BF deals etc.). This could get them to make the jump early. Which I'm sure is the mindset at MS. I expect MS will kill production on Xbone (might have done so already, LOL) as soon as the new console launches, so keeping a lower cost sku could help there to.

 

[Hawking Radiation]

People saying 9.2TF is locked on.

We have had a handful of industry sources and insiders pegging PS5 as being faster of the two.

Klee confirmed both are above 10TF.

We also know there were 2 versions of dev kits and the one around the Wired article was the reported low speed version.

How likely is it that the recent AMD testd were of the low speed PS5 dev kit and not the current one?

How can it be that both are reportedly basically power twins and very close to each other that performance wont matter?
Yet people on here and Era are conceding PS5 being at 9.2TF for retail?

It makes no sense to me.

 

[vivftp]

People saying 9.2TF is locked on.

We have had a handful of industry sources and insiders pegging PS5 as being faster of the two.

Klee confirmed both are above 10TF.

We also know there were 2 versions of dev kits and the one around the Wired article was the reported low speed version.

How likely is it that the recent AMD testd were of the low speed PS5 dev kit and not the current one?

How can it be that both are reportedly basically power twins and very close to each other that performance wont matter?
Yet people on here and Era are conceding PS5 being at 9.2TF for retail?

It makes no sense to me.

That's the theory I'm running with

 

[vpance]

Well you can also say that if the PS5 is 400, you might as well pay an extra 100 and get the better Xbox. It all depends on how people see things doesn’t it.

I think another reason for Lockhar’s existence is xcloud. They can use that in the servers because they are primarily targeting phones and tablets where 4K is largely pointless. While they are at it they can do a cheap retail version at the same time to cover more of the market.

I’m just guessing for fun. But one thing for certain is that those Codenames and chip tests are heavily tied to both PlayStation and Xbox hardware. That is undeniable.

Don't forget that XCloud tech needs to run Sony games as well. They're partners now. Wouldn't be surprised if all this data is just tests for their joint streaming platform.

 

[demigod]

Some of the headcanon in here is pretty amusing to read.

I'm sitting back here enjoying the show.

 

[Lort]

I can see it selling well, I've bought the last Xbox consoles straight out of the gate, so $500 is a typical price for me. But a lot of buyers wait on next gen until prices decrease (those first good BF deals etc.). This could get them to make the jump early. Which I'm sure is the mindset at MS. I expect MS will kill production on Xbone (might have done so already, LOL) as soon as the new console launches, so keeping a lower cost sku could help there to.

Its enough to play xbox one x games at 4k 30 fps and next gen games at 1440p or 1080p .. at 60 fps.

The xbox one x will be phased out and we will end up with
Xbox stream (xbox one s ) $149 streaming console
Xbox series x $299
Xbox series x Pro $499

 

[Hawking Radiation]

I'm sitting back here enjoying the show.

You know something right? I mean you have inside info on both machines?
Dont worry I'm not asking you to put out any information on them. I'm happy to wait.

 

[Hawking Radiation]

That's the theory I'm running with

Yep. Seen your post and agree with it.

Just confused why some are just ready to accept 9.2TF PS5 when all other information from people in the know point to something alot more potent than that.

 

[vivftp]

Thoughts?

 

[Dee Dah Dave]

Zen 3 and RDNA2 are on 7nm+

PS / Xbox will be custom RDNA on 7nm

 

[ANIMAL1975]

I think you do not understand the actual chip design process and how much time and money it takes to put such a big and fast chip on the market.

Look...they wont have specific chips just for dev kits. If that was the case, they could have used actual 5700XT GPUs and get on with it. But they didnt, they have been working on Ariel/Oberon for years. It is what it is. Oberon latest revision is from May, which tells you all you need to know really.

It's a devkit, if they can clock it up to 2GHz to squeeze more power out of it then why not? The only objective is to squeeze more power out of this temporary early devkit until the next revision is ready. It's in a huge tower where it can be cooled out the ass and things like noise aren't a concern like in a typical home setup.

Haven't you seen already you wasting your time? If you say a cow is a cow he will say you're wrong it's a flute...
... all Shakespeare names and all related

 

[Rightisright]

All evidence? Please. This one Github leak points to that. All the evidence we have up until this point is to the contrary. That includes months of insiders and journalists saying they're hearing about a PS5 that's roughly on-par with Scarlett and some indicating it might be a bit higher. That's including Jason Schreier, Matt Piscatella, Tom Warren and Kleegamefan.

The only other things we ever heard to indicate a 9.2 TF console was someone mentioning they heard that figure at GDC, and that's easily explained by the fact that only the low speed devkits would have been handed out up until that point with the V shaped devkits going out afterwards.

Anyways, I fully agree we now need to wait for more information. I do think this is a very solid theory though and I'll choose to run with it until we hear otherwise.

Actual evidence is AMD documentation showing clock speeds, CU counts, bandwidth etc.
Heresay is someone saying their friend is an insider and he says they are similar power, but I'm in Alaska chopping down trees, so I cant comment on all the AMD evidence that contradicts what I have said.
Have you noticed non of these so called insiders with spec sheets have said anything to go against the leaks? None of them? Wonder why?

And let's see what people have been saying about these consoles.

Tom Warren
" hardly anyone has them, and they’re nowhere near final. That’s why you’re hearing a lot of nonsense about them not being powerful etc. Microsoft wants to surprise Sony this time"
Doesn't sound like Sony is more powerful from that tweet.

Albert Penello says that he predicts the PS5 to be between 8-9.2 tflops. He has said this for awhile.

So show me where Matt or Klee etc have come out to dispute the leaks? They haven't, they have gone underground.

 

[Fafalada]

Problem is they have gone with 256bit bus and with fastest available memory next year these memory modules would still not bring enough BW for 10TF.

How exactly do you compute 'enough BW for X TF'(and please don't say it's based off of PC title benchmarks) - because unlike the old-style 'how do you hit maximum ROP utilization' I'm aware of no such straightforward metric for compute. I'm also asking because whatever people data-mined here claims the 12TF box is supposed to hit around 550GB/s which according to your observation would be a significant shortfall as well?

 

[vivftp]

Yep. Seen your post and agree with it.

Just confused why some are just ready to accept 9.2TF PS5 when all other information from people in the know point to something alot more potent than that.

I don't think there's any one answer. For some people, it might be a console war thing. For others there are certain narratives they'd prefer to believe based on the available data (ie. the stuff about Jim Ryan wanting to transition people over to PS5 at record pace, so people think that must mean it's $399). Or others might be taking it more seriously than other rumors because these are actual numbers from an official source and they find that more believable than rumors.

Whatever it is, I do hope this theory about it being the low speed devkit is accurate as I would like a more powerful PS5 personally

 

[HeisenbergFX4]

Actual evidence is AMD documentation showing clock speeds, CU counts, bandwidth etc.
Heresay is someone saying their friend is an insider and he says they are similar power, but I'm in Alaska chopping down trees, so I cant comment on all the AMD evidence that contradicts what I have said.
Have you noticed non of these so called insiders with spec sheets have said anything to go against the leaks? None of them? Wonder why?

And let's see what people have been saying about these consoles.

Tom Warren
" hardly anyone has them, and they’re nowhere near final. That’s why you’re hearing a lot of nonsense about them not being powerful etc. Microsoft wants to surprise Sony this time"
Doesn't sound like Sony is more powerful from that tweet.

Albert Penello says that he predicts the PS5 to be between 8-9.2 tflops. He has said this for awhile.

So show me where Matt or Klee etc have come out to dispute the leaks? They haven't, they have gone underground.

Klee has said he is in Alaska working 12 hour shifts I dont think his priorities include reassuring people his mind hasnt changed.

He did say at one time the 29th, today, was the day he gets to head home so maybe he will post again soon.

 

[demigod]

You know something right? I mean you have inside info on both machines?
Dont worry I'm not asking you to put out any information on them. I'm happy to wait.

Nope I don't. Regardless of PS5's power I'm still getting it. Some people shilling hard in here.

 

[Hawking Radiation]

Actual evidence is AMD documentation showing clock speeds, CU counts, bandwidth etc.
Heresay is someone saying their friend is an insider and he says they are similar power, but I'm in Alaska chopping down trees, so I cant comment on all the AMD evidence that contradicts what I have said.
Have you noticed non of these so called insiders with spec sheets have said anything to go against the leaks? None of them? Wonder why?

And let's see what people have been saying about these consoles.

Tom Warren
" hardly anyone has them, and they’re nowhere near final. That’s why you’re hearing a lot of nonsense about them not being powerful etc. Microsoft wants to surprise Sony this time"
Doesn't sound like Sony is more powerful from that tweet.

Albert Penello says that he predicts the PS5 to be between 8-9.2 tflops. He has said this for awhile.

So show me where Matt or Klee etc have come out to dispute the leaks? They haven't, they have gone underground.

They don't have to do anything. Remember they have sources that they would like to protect. Imagine coming to a gaming forum to run your mouth on specs and put in jeopardy the jobs of the industry insiders from whom you got the information from. This is playing with the livelihood of another person.

To be fair they have said enough already and its a very short time until we know the outcome of this debate.

Albert 'predicts' is right. Key word here. He doesn't know anything and he no longer works for Microsoft.

 

[chemical boy]

Thoughts?

That tastes the same as the rest but seeks attention because a few days ago he wrote that he did not know the hardware of the PS5 and now instead says that Xbox is more powerful.

 

[demigod]

That tastes the same as the rest but seeks attention because a few days ago he wrote that he did not know the hardware of the PS5 and now instead says that Xbox is more powerful.

Yup. He was boasting about Scarlett's power until E3, then went into hiding. After TGA he showed his face again. They have no info on PS5 other than _rogame which was old stuff.

 

[Hawking Radiation]

Nope I don't. Regardless of PS5's power I'm still getting it. Some people shilling hard in here.

I share the sentiment. I have invested too much money in the PlayStation ecosystem to pick up an Xbox Series X.

100 + games and counting. My internal 4TB SSD and system 1TB can't keep up. Have had to delete games to make the most played ones fit LOL.

Both Series X and PS5 will be great and most are surely underestimating the jump in storage, CPU and raytracing.

I honestly believe we are going to see some shit that none of us are prepared for. Just looking at Modern Warfare, Uncharted 4 and Gears 5 running on Xbox and PS4 with laptop grade CPU's have me foaming at the mouth at the thought of devs having damn near 10TF's and desktop class CPU's to play with.

I mean FUCK. I can't wait. My Pro already makes a helluva noise and i'm ready to retire her to the closet in favour of my shiny new PS5 lol

 

[R600]

How exactly do you compute 'enough BW for X TF'(and please don't say it's based off of PC title benchmarks) - because unlike the old-style 'how do you hit maximum ROP utilization' I'm aware of no such straightforward metric for compute. I'm also asking because whatever people data-mined here claims the 12TF box is supposed to hit around 550GB/s which according to your observation would be a significant shortfall as well?

Well it is based on 448GB/s in 5700XT, which would be a similar card.

I said I doubt it, because of already very, very high clocks. In any case, what works for sony works for MS, but I guess they will fall into similar terriory regarding TF per GB/s of BW.

 

[Rightisright]

With regards to these ERA insiders, all they did was just give motherhood statements such as "They are pretty close in power", "not much between them", "maybe a slight edge to PS5, but that could change before release".
Not one of them gave a single spec of speed or CU count or anything. Nothing at all. Any flog could have said the same thing as they did because that was pretty much what most people thought it would be this gen. This was because they would be using the exact same CPU and the same Navi tech. Brilliant deductions. Not.

So when we have ACTUAL leaks that come out in Feb this year that showed the exact Xbox specs for Lockhart and Anaconda, people dismiss this for the unsubstantiated basic rumours put out by the Era insiders.

However when it turns out that quite possibly PS5 isn't as powerful as the Xbox from AMD data, then these guys disappear.

 

[Proelite]

Zen 3 and RDNA2 are on 7nm+

PS / Xbox will be custom RDNA on 7nm

Bingo.

I thought there was a >50% chance that Xsx was an 7nm euv, since it has 56CUS.

But if the SOC was the upper limit of my die size estimate, 370mm2-380mm2, it can have 56CUs active on 7nm.

 

[R600]

Bingo.

I thought there was a >50% chance that Xsx was an 7nm euv, since it has 56CUS.

But if the SOC was the upper limit of my die size estimate, 370mm2-380mm2, it can have 56CUs active on 7nm.

Judging by Flute benchmark, they cut the Zen2 cache by fair amount.

Cutting 3/4 of L3 cache will probably result in ~35mm² less.

 

[XcloudTimdog]

Bingo.

I thought there was a >50% chance that Xsx was an 7nm euv, since it has 56CUS.

But if the SOC was the upper limit of my die size estimate, 370mm2-380mm2, it can have 56CUs active on 7nm.

Hmmq confirmed to be 7nm , looks to be navi12 56cu, custom design.

 

[chemical boy]

Now he suddenly says he has PS5 information ... what a coincidence!

 

[HeisenbergFX4]

With regards to these ERA insiders, all they did was just give motherhood statements such as "They are pretty close in power", "not much between them", "maybe a slight edge to PS5, but that could change before release".
Not one of them gave a single spec of speed or CU count or anything. Nothing at all. Any flog could have said the same thing as they did because that was pretty much what most people thought it would be this gen. This was because they would be using the exact same CPU and the same Navi tech. Brilliant deductions. Not.

So when we have ACTUAL leaks that come out in Feb this year that showed the exact Xbox specs for Lockhart and Anaconda, people dismiss this for the unsubstantiated basic rumours put out by the Era insiders.

However when it turns out that quite possibly PS5 isn't as powerful as the Xbox from AMD data, then these guys disappear.

They will be close but Xbox will have a slight advantage.

 

[Hawking Radiation]

I don't think there's any one answer. For some people, it might be a console war thing. For others there are certain narratives they'd prefer to believe based on the available data (ie. the stuff about Jim Ryan wanting to transition people over to PS5 at record pace, so people think that must mean it's $399). Or others might be taking it more seriously than other rumors because these are actual numbers from an official source and they find that more believable than rumors.

Whatever it is, I do hope this theory about it being the low speed devkit is accurate as I would like a more powerful PS5 personally

Sure i'd love a powerful PS5 too. I mean as powerful as possible but either way i'm still pre-ordering one.

I know Sony will fix the issues of the PS4.

- Poor dualshock battery life
- Loud system noise (Modern Warfare has me legit worried my Pro will break at some point)
- Cheap build quality (they dropped the ball on the Pro here)
- Better and more responsive PSN interface

Of the above I am certain they will fix all of the issues.

Either way like i said before, both Series X and PS5 will be amazing. Its going to be so refreshing to just focus on the games and not counting the blades of grass in one version compared to the other.
That red dead digital foundry episode still gives me nightmares LOL

 

[Rightisright]

Klee has said he is in Alaska working 12 hour shifts I dont think his priorities include reassuring people his mind hasnt changed.

He did say at one time the 29th, today, was the day he gets to head home so maybe he will post again soon.

And yet he has been logging into Era quite frequently but just not commenting. He has that spec sheet right? And if he doesn't have it on him, its not hard to remember how many tflops a console has is it now?
And you would have noticed that he never mentioned one spec. Not one. Just a bland comment about them being similar spec, maybe PS5 with the edge, but that could change. And the time he was saying this was the same time AMD was doing benchmarks on a 36cu, 2ghz GPU card. Funny how he got the updated PS GPU while stupud AMD were still benchmarking the old one......

 

[Dee Dah Dave]

I’m sure they would have liked to be on 7nm+, but these things are designed well in advance. They can only use what’s available at the time of development, maybe slightly ahead. 7nm+ is just out of reach to 100% guarantee a holiday 2020 release, such is life.

Also a good lesson in how far ahead these products need to be finished before they actually launch, there has to be a cutoff point somewhere. Always.

 

[demigod]

Hmmq confirmed to be 7nm , looks to be navi12 56cu, custom design.

I thought you, proclaimed insider, had info on the X, why you asking him?

 

[Fafalada]

Well it is based on 448GB/s in 5700XT, which would be a similar card.

True but it doesn't exactly run the same software.
While more BW never hurts, I do think we're getting to area where ROPs don't have a constant shortfall anymore at target resolutions (something that still punished current consoles, especially the Pro) so it may not be such a bad place. Although RT will change things again - so harder to predict with that, but we have literally no precedent to work from here.

 

[TLZ]

I'm sitting back here enjoying the show.

Thoughts?

So, are people forcing a certain narrative going to quote this, like the other quotes they posted here? Or are they leaving it behind because it doesn't fit? Or are they going to say he doesn't know, even though he was quoted previously as proof of being knowledgeable? Xbox SX was rDNA2 all along, wasn't it?

 

[HeisenbergFX4]

And yet he has been logging into Era quite frequently but just not commenting. He has that spec sheet right? And if he doesn't have it on him, its not hard to remember how many tflops a console has is it now?
And you would have noticed that he never mentioned one spec. Not one. Just a bland comment about them being similar spec, maybe PS5 with the edge, but that could change. And the time he was saying this was the same time AMD was doing benchmarks on a 36cu, 2ghz GPU card. Funny how he got the updated PS GPU while stupud AMD were still benchmarking the old one......

So lets say he does have said spec sheet.

Is he supposed to give everything he has to ERA to assure everyone anytime a supposed leak happens?

He has repeated several times they were close in power why does he need to post that every 5 pages?

 

[TLZ]

That tastes the same as the rest but seeks attention because a few days ago he wrote that he did not know the hardware of the PS5 and now instead says that Xbox is more powerful.

Exactly.

 

[vivftp]

Actual evidence is AMD documentation showing clock speeds, CU counts, bandwidth etc.
Heresay is someone saying their friend is an insider and he says they are similar power, but I'm in Alaska chopping down trees, so I cant comment on all the AMD evidence that contradicts what I have said.
Have you noticed non of these so called insiders with spec sheets have said anything to go against the leaks? None of them? Wonder why?

And let's see what people have been saying about these consoles.

Tom Warren
" hardly anyone has them, and they’re nowhere near final. That’s why you’re hearing a lot of nonsense about them not being powerful etc. Microsoft wants to surprise Sony this time"
Doesn't sound like Sony is more powerful from that tweet.

Albert Penello says that he predicts the PS5 to be between 8-9.2 tflops. He has said this for awhile.

So show me where Matt or Klee etc have come out to dispute the leaks? They haven't, they have gone underground.

Wow, where to start...

You do realize that I'm not disputing the info in the Github leak, right? I believe it's 100% true. What I'm disputing is the interpretation of what that data means. Many took it to assume it was a devkit that closely matched what the final PS5 will have. I'm taking it to mean that it's representative of the "early low-speed devkit" that we know Cerny showed off to Wired just a couple months before these tests were run. I've stated my reasons for believing this to be the case.

Regarding Tom Warren, and his comment about MS wanting to surprise Sony, you do realize he said this just after they did a surprise reveal of the XSX at the Game Awards, right?

Also, did you know at E3 he reported the rumored devkit for the PS5 was 13 TF,? This would have been after more devs started to receive the V shaped devkit:



Albert Penello stated he was guessing ~8 TF for the PS5 based on his assumption that they were going for a $399 price point. He doesn't have any particular insight into what Sony is doing and is fully open to being wrong:

As for Matt and Klee, I was talking about their discussions about the specs between XSX and PS5 over the past several months. No insider is touching the Github leak with a 10 foot pole because it's a hot potato right now and no one can blame them for not wanting to subject themselves to that nonsense. They've said what they have to say and it all fits beautifully in the theory I'm discussing.

 

[Lort]

Thoughts?

This is very interesting ( esp his other posts).

His basically saying neither are technically rdna2.. but both are custom with most of the same architecture.

The xbox 360 had integrated pixel and vertex shaders but wasnt actually a specific derivative of any amd graphics card.

I think this might be on the money..

Basically this would be saying all leaks are true .. ps has some type of ray tracing but its different to xbox ( or was not enabled at that time)..arden is def Xbox and its drivers detect it and vrs.

Xbox is ahead on tflops but has custom optimizations to make it better than it would else have been, he dosent know ps5 ray tracing or other optimizations are implemented but thinks it wont be enough to overcome xbox tf advantage.

This would explain a lot of things including why the numbering and versioning is so odd ( the ps and xbox chips were potentially in testing before any amd rdna2 chip)... why they are potentially on 7nm not 7nm+ and why xbox refuses to confirm if tf are gcn or navi .. because technically its neither.

 

[BattleScar]

Hmmq confirmed to be 7nm , looks to be navi12 56cu, custom design.

????????
Both Rogame and Komachi's information from the GitHub repo/whatever datamining they did has more or less unequivocally stated that Navi 12 is 40CU with HBM2. It apparently has some hardware fixes, but is essentially Navi 10 but with a different memory controller.

So no, its not Navi 12.

 

[R600]

????????
Both Rogame and Komachi's information from the GitHub repo/whatever datamining they did has more or less unequivocally stated that Navi 12 is 40CU with HBM2. It apparently has some hardware fixes, but is essentially Navi 10 but with a different memory controller.

So no, its not Navi 12.

Both are custom RDNA derivatives for christ sake guys.

Its how custom APUs for consoles have always worked. Thats why PS4 Pro is Polaris based with Vega features. They CAN use and abuse anything AMD has on roadmap and put it on chip.

You dont actually take PC card and make an SOC for console out of it.

Oberon regression didnt show RT because it wasnt even a point of test, not because RT is not there.

 

[Clintizzle]

Fuck it, i'm skipping this gen.

 

[R600]

Anyone remembers this gem? Xbox 360 spec leak in April 2004, 1.5years before it was released.

Xbox 360 tech a decade old - Xenon System Diagram leaked April 2004

I recall the diagram being called fake by some of the press and individuals. GameSpot Rumor Control April 30, 2004 RUMOR #2: A technical diagram of the Xbox Next has been leaked. Source: Chinese hardware site GZeasy.com. The official story: "Microsoft does not comment on rumor...

forum.beyond3d.com

You can see how far they are designed and locked from this. Some things can change (mem amount and clocks, but chip design no).

 

[CyberPanda]

Anyone remembers this gem? Xbox 360 spec leak in April 2004, 1.5years before it was released.

Xbox 360 tech a decade old - Xenon System Diagram leaked April 2004

I recall the diagram being called fake by some of the press and individuals. GameSpot Rumor Control April 30, 2004 RUMOR #2: A technical diagram of the Xbox Next has been leaked. Source: Chinese hardware site GZeasy.com. The official story: "Microsoft does not comment on rumor...

forum.beyond3d.com

You can see how far they are designed and locked from this. Some things can change (mem amount and clocks, but chip design no).

Yep. Well said

 

[Rightisright]

So lets say he does have said spec sheet.

Is he supposed to give everything he has to ERA to assure everyone anytime a supposed leak happens?

He has repeated several times they were close in power why does he need to post that every 5 pages?

He hasn't posted ANY specs. Not even a range. Nothing.
If you are gonna go public and proclaim that you have inside knowedge of the console specs, and take the fame of everyone sayong "Klee said this" "Klee said that" then its pretty reasonable for you to have to pony up with some facts.

 

[chemical boy]

I think that Sony are really happy, some people have lowered the PS5 specifications so much that when they are announced they will really be a blow of effect. If the idea was to damage the PS5 image, if the console finally has 13 tflops, the impact will be even greater than what was expected a month ago. Lowering expectations has come great.

 

[XcloudTimdog]

I thought you, proclaimed insider, had info on the X, why you asking him?

I know its 12 rdna tflops. How it gets there is the question.

 

[Gamernyc78]

I think that Sony are really happy, some people have lowered the PS5 specifications so much that when they are announced they will really be a blow of effect. If the idea was to damage the PS5 image, if the console finally has 13 tflops, the impact will be even greater than what was expected a month ago. Lowering expectations has come great.

Cerny and Shu must be laughing their asses off with the mad scientist vibe going on.

 

[Dee Dah Dave]

Anyone remembers this gem? Xbox 360 spec leak in April 2004, 1.5years before it was released.

Xbox 360 tech a decade old - Xenon System Diagram leaked April 2004

I recall the diagram being called fake by some of the press and individuals. GameSpot Rumor Control April 30, 2004 RUMOR #2: A technical diagram of the Xbox Next has been leaked. Source: Chinese hardware site GZeasy.com. The official story: "Microsoft does not comment on rumor...

forum.beyond3d.com

You can see how far they are designed and locked from this. Some things can change (mem amount and clocks, but chip design no).

^Exactly. So the PS5 chip is either -

a) real

or

b) there’s another one hiding somewhere

 

[XcloudTimdog]

????????
Both Rogame and Komachi's information from the GitHub repo/whatever datamining they did has more or less unequivocally stated that Navi 12 is 40CU with HBM2. It apparently has some hardware fixes, but is essentially Navi 10 but with a different memory controller.

So no, its not Navi 12.

Yeah thats curious, i asked him for a follow up and he didnt answer. He brought up navi12