• 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.

hardseppo

Member
PSA: All insiders are fake. Only believe in official statements from Jim and Phil or press-releases.

900 pages of this thread would probably not be there, if it wasn't for those attention-seeking "insiders". I believed in them and personally now I have enough.

Actually I am also mad at my myself for believing all this crap. Where is the event announcement?

PS: I personally think that PS5-Sphere design is sexy. I want it. There, I said it. At least I hope that is true.
 
Last edited:

SlimySnake

Flashless at the Golden Globes
OK, guys, can anyone tell what does “1st hand info” or 2nd hand info or 3rd hand? Since I’m a foreign guy I do not understand what that means.
First hand info would be from a dev themselves posting here who have access to devkits and specs.

Second hand is basically your insiders like osiris and kleegamefan or reporters.

Third hand is Tommy Fisher's hand jerking us off.
 

ethomaz

Banned
CU amount was determined by the number of stream processors which was 3584. Navi has 64 SP per CU, that's 56CUs.
Yeap but we don't know how many CUs Xbox Series X have.
Neither the clock of the GPU for both Xbox Series X or GitHub.

Using memory clock to calc with the 56CUs the TFs is so wrong.

So Xbox Series X revealed specs did not confirm anything on GitHub because there is no related specs to compare.
 
Last edited:

Tenka Musou

Banned
But you described a sizeable part of the demo you saw. That already is enough for higher ups to know you are talking about THEIR game. so they would know of a leak
Not only that, but he specified the role of the person who was showing him the demo (think he said it was the art lead or something).
 
The range is very wide. You could say scientifically that it can start between 9 and 15. And it can end between 19 and 21. Although it usually goes between 10 and 17. I personally think I started it around 11. I would like to add that the first menstruation usually occurs between 12-13, while the first ejaculation usually appears at 13. It is in the only average that we agree men and women. All this is real scientific data that subscribe to a trend, but obviously it is not a fixed norm, it is variable according to the person, but it would become a good reference.
[...]

The first devkit he used was at around 11;
one device now between 12 and 13
one device now at around 13

??
 

MadAnon

Member
He seemed to take the bitch way out, so explain it.
Here you go...

 

DeepEnigma

Gold Member
Here you go...


How convenient. :messenger_weary::messenger_ok:

Especially the backup dancers, lol.
 
Last edited:

MadAnon

Member
Yeap but we don't know how many CUs Xbox Series X have.
Neither the clock of the GPU for both Xbox Series X or GitHub.

Using memory clock to calc with the 56CUs the TFs is so wrong.

So Xbox Series X revealed specs did not confirm anything on GitHub because there is no related specs to compare.
That's fair but if we take some of those Navi efficiency curves into account. The sweet spot frequency with 56CUs would equate ~12TF.
 

Mr Moose

Member
Here you go...

People 100% believe it even though it contains errors.
 

ethomaz

Banned
Here you go...

That is not a explanation but a guess.
 

ethomaz

Banned
That's fair but if we take some of those Navi efficiency curves into account. The sweet spot frequency with 56CUs would equate ~12TF.
That is true... if you assume the GitHub is right and Xbox One X 12TFs you can easily find the clock.
I'm just saying the 12TFs doesn't confirme the GitHub data.
 
Last edited:

OsirisBlack

Banned
Not only that, but he specified the role of the person who was showing him the demo (think he said it was the art lead or something).
Misdirection:
the action or process of directing someone to the wrong place or in the wrong direction.
"the deliberate misdirection that had put me off the track"
Side note that studio has since closed to all media and tightened up ship. You can verify that on their press page if you can figure out which studio it was.

What I sent the mod was not tied to that company (based in California) but my exact location, (Not in California) who I was with and what I was doing. It is too exact to share as I also sent phone records. So no that will not be made public.
 

ethomaz

Banned
and does that make sense to you?

how can something more powerful offer worse performance?

its clear we need to learn from today and ignore everyone's riddles.
Performance is not related to GPU power only...

For example memory bandwidth play a big whole in performance so a 15TFs GPU can perform worst than a 14TFs just because the 14TFs has better memory bandwidth.
Now take in account CPU, SSD, memory, etc etc etc.

Plus the software side... the SDK, the API, etc.

It is perfectly possible to have more TFs and have less performance.

PS. I'm not saying PS5 have better performance with lower TFs... I'm just saying it is possible to that happen.
 
Last edited:

JMB88

Neo Member
I think we have been given too much relevant information. Knowing that power is similary between the 2 things is enough.

Knowing absolutely all the specifications and features would take away all the surprise factor from the future announcement. And what's so funny about that?

And Tommy was predictable. People's problem is that they believe absolutely everything when they tell them what they want to hear.

That's why politicians are so successful...
 
Last edited:

DJ12

Member
You guys are going IN on Tommy Fisher Tommy Fisher 😂😂

It still pains me that he was a fake insider all along 😭😭😭
Come on now, while it's not looking good for Tommy, Heinsenberg was told by a SIE employee two or three weeks ago that the PS5 was 9.2 TFs, when it's 100% certain it isn't as BGs has stated many times now.

Tommy might be an unwitting partner in some more Sony based trolling.

I think at the very least, should he choice to, he should stick around until the reveal, if anything he's said is true, then you cannot really argue he's a legit insider in my opinion.
 
Actually what I said was.

"I'm going to help dig yourself out of this hole, remember when someone (I have no idea who) said they heard the machines were 11.6 and 12.4 sometime last year? That number was SPOT ON at that time with PSV having a slight lead but also having heat and sound issues. (Heat and sound are no longer issues) MS announces 12 TF hey look a .4 increase not bad at all they made it. BOTH are > 10 (said this last year) with PSV still having a slight advantage. (Said this last year also.) Both were aiming for a very high number (13) and from what I can gather one of them may have actually hit that number. (Its not MS) Either way BOTH consoles are stupidly powerful and still very close. 12 vs 12.4 or 13.2 (Within 1TF of the numbers with a . up in both cases not down) isn't bad at all."

That's me doubling down on 12.4 (which i said from the get go) or Higher. Nothing lower.
Thanks for this OsirisBlack OsirisBlack .
This thread moves way too fast. Could a kind individual/ Mod please add all that we have learned today and from BGs into the OP? (I apologize if it has already been done) Thanks.
 

SlimySnake

Flashless at the Golden Globes
Here you go...

this theory never made sense to me. Why is a fucking intern grabbing data for both Arden and Oberon? There is NO way AMD is allowing that kind of cross talk between engineers, let alone giving access to fucking interns.

If it's a script pulling data then the script must have known the names of PS5 chips AND xbox chips. Why is he compiling all that stuff in one file? Who is he doing that for? the teams are separate. this is confidential stuff. if i was Sony, i would sue AMD for negligent behavior and charge this intern with corporate espionage.
 
Last edited:

FeiRR

Banned
I know it's all fun but you all should realise the people involved aren't going to risk their careers for the sake of your entertainment. If we don't get any straight answers, just deal with it and don't worry. Your favourite brand operator will inform you beforehand about all the intricacies of the hardware you want to buy well ahead of the launch date.

Because it's your money into their pockets.
 

SlimySnake

Flashless at the Golden Globes
Performance is not related to GPU power only...

For example memory bandwidth play a big whole in performance so a 15TFs GPU can perform worst than a 14TFs just because the 14TFs has better memory bandwidth.
Now take in account CPU, SSD, memory, etc etc etc.

Plus the software side... the SDK, the API, etc.

It is perfectly possible to have more TFs and have less performance.

PS. I'm not saying PS5 have better performance with lower TFs... I'm just saying it is possible to that happen.
then why not say PS5 is more powerful? specs arent just GPU tflops. if the PS5 is doing everything else better like memory bandwidth, ssd speed and cpu speed then PS5 is the more powerful console.
 

icerock

Member
Has anyone ever been able to explain this?

Most common explanation has been that, someone made an error.

Besides, while I don't think it is wise to hand wave the Github data, one shouldn't treat it as gospel either. To this date, no one can point out the total CU count on Oberon (iGPU) that was being tested there, unlike Arden. We know the 18WGP was marked as "Full chip result", which would imply 36 active CUs but how many were disabled?

There is no product from AMD out there which has 36/36 active CUs, so it had to be using an iGPU which has some redundant CUs. Most people assumed it was a 5700 chip where 4CUs had been disabled, and logically it makes sense too, given the yields, and previous iteration we saw sharing the same codename. But, at the the same time, who can confidently say, that it wasn't a bigger NAVI chip? It's not like, there weren't in existence at the time, given the same repo gave us details about Arden.

The data is very informative, but it doesn't give us full picture. Those who are saying it does, are just trying to connect different dots and filling their own ideas when there is a gap.
 
Last edited:

SleepDoctor

Banned
People 100% believe it even though it contains errors.


People were just praising "Tommy" like he was a god. Even after the bullshit Feb reveal and not self banning himself.

Like I said tho, fanboys will go with anything that fits their narrative. I could probably make alt and say a couple things like "ps5 20tf lolz xbots fail" and a few "special" people here would run with it lol.
 
Come on now, while it's not looking good for Tommy, Heinsenberg was told by a SIE employee two or three weeks ago that the PS5 was 9.2 TFs, when it's 100% certain it isn't as BGs has stated many times now.

Tommy might be an unwitting partner in some more Sony based trolling.

I think at the very least, should he choice to, he should stick around until the reveal, if anything he's said is true, then you cannot really argue he's a legit insider in my opinion.

Interesting. Wait. Do you have the quote from Heisenberg that a Sony employee told him PS5 was 9.2 TF?
 
Last edited:

DJ12

Member
I'm looking forward to what AMD reveal on Thursday, if it goes ahead 🤔
LOL, 5 hours of investor related powerpoints aimed at the guests on site.

Good luck with that pal. Investors don't care about how pretty something looks, they care how much money they are going to make from it.

As an example of what to expect, please look here:


Thats a previous event, this is the current one:


Please note only a speaker symbol near the webcast link, they aren't even broadcasting a video of the event.
 
Last edited:

DeepEnigma

Gold Member
Most common explanation has been that, someone made an error.

Besides, while I don't think it is wise to hand wave the Github data, one shouldn't treat it as gospel either. To this date, no one can point out the total CU count on Oberon (iGPU) that was being tested there, unlike Arden. We know the 18WGP was marked as "Full chip result", which would imply 36 active CUs but how many were disabled?

There is no product from AMD out there which has 36/36 active CUs, so it had to be using an iGPU which has some redundant CUs. Most people assumed it was a 5700 chip where 4CUs had been disabled, and logically it makes sense too, given the yields, and previous iteration we saw sharing the same codename. But, at the the same time, who can confidently say, that it wasn't a bigger NAVI chip? It's not like, there weren't in existence at the time, given the same repo gave us details about Arden.

The data is very informative, but it doesn't give us full picture. Those who are saying it does, are just trying to connect different dots and filling their own ideas when there is a gap.

The most logical explanation I’ve heard yet.
 
Last edited:

SlimySnake

Flashless at the Golden Globes
Come on now, while it's not looking good for Tommy, Heinsenberg was told by a SIE employee two or three weeks ago that the PS5 was 9.2 TFs, when it's 100% certain it isn't as BGs has stated many times now.

Tommy might be an unwitting partner in some more Sony based trolling.

I think at the very least, should he choice to, he should stick around until the reveal, if anything he's said is true, then you cannot really argue he's a legit insider in my opinion.
to be clear, heisenberg was told ps5 was 9.5 tflops and we dont know if his source is from SIE.
 

geordiemp

Member
this theory never made sense to me. Why is a fucking intern grabbing data for both Arden and Oberon? There is NO way AMD is allowing that kind of cross talk between engineers, let alone giving access to fucking interns.

If it's a script pulling data then the script must have known the names of PS5 chips AND xbox chips. Why is he compiling all that stuff in one file? Who is he doing that for? the teams are separate. this is confidential stuff. if i was Sony, i would sue AMD for negligent behavior and charge this intern with corporate espionage.

Agree its crazy an intern has all the data for 2 new next gen consoles in one file,. and he stores it where it can be seen, its crazy.
 

nikolino840

Member
LOL, 5 hours of investor related powerpoints aimed at the guests on site.

Good luck with that pal. Investors don't care about how pretty something looks, they care how much money they are going to make from it.

As an example of what to expect, please look here:


Thats a previous event, this is the current one:


Please note only a speaker symbol near the webcast link, they aren't even broadcasting a video of the event.


 
Last edited:
Status
Not open for further replies.
Top Bottom