• Light/Dark
  • Ever wanted an RSS feed of all your favorite gaming news sites? Go check out our new Gaming Headlines feed! Read more about it here.
  • We have made minor adjustments to how the search bar works on ResetEra. You can read about the changes here.

PS5 Speculation |OT12| - Aw hell, Transistor's running this again?

How hyped are you?

  • A little hyped

    Votes: 98 15.7%

  • A lot of hyped

    Votes: 50 8.0%

  • WALNUUUUUTSSSSS

    Votes: 222 35.5%

  • Hyped enough to eat this whole bag of walnuts

    Votes: 63 10.1%

  • Hyped enough to bite this moose

    Votes: 37 5.9%

  • Hyped enough to scramble a dozen eggs

    Votes: 39 6.2%

  • Hyped enough to be even more hyped, like, cyclical or something

    Votes: 116 18.6%
  • Total voters
    625
Status
Not open for further replies.
Reader mode
modiz

modiz

Member
Oct 8, 2018
17,841
RoninStrife said:
I see. Thanks. Seeing guys throw 13TF as a possible spec .. very exciting times. I will be happy with 12.. or more.
Click to expand...
Click to shrink...
With those IPC gains, i'll be very happy with 10TF, that should be already enough to do 4k 60 on every current gen game, which would mean good increase to graphical complexity while staying at a high resolution, so honestly I am not really bothered by any TF number now.

AegonSnake said:
The curious thing to me is that MS codeveloped RDNA 2.0 with AMD. Does sony get to use on rdna 2.0? its funny how we thought navi was made for sony but it seems it was MS who was working on the tech thats going in next gen consoles or at least one next gen console.

it will be curious to see what makes Sony's custom SOC different in case MS had some kind of exclusivity on rdna 2.0. are we looking at the same perf/watts ratio? is RT the same? wtf were the github leaks about if those 9.2 tflops are actually supposed to mimic 9.2 rdna 2.0 tflops? is sony like 6 rdna 2.0 tflops?
Click to expand...
Click to shrink...
they co developed the software API for DXR, not the actual hardware architecture.
 
Shambala

Shambala

Banned
Oct 26, 2017
1,537
I can wait to call out certain people once the full specs are revealed. There sure are some know it all posters on this website 😂
 
MrKlaw

MrKlaw

Member
Oct 25, 2017
33,056
Marble said:
Yes, but theoretically: is 9TF with RDNA 2 logical, or will it most likely be more around 12TF?
Click to expand...
Click to shrink...

if its 36CU at 2Ghz then it's still 9.2TF. Just that the power efficiency gains would mean that is more reasonable and not a complete power hog, and therefore easier to cool, smaller power supply etc
 
Doctor Avatar

Doctor Avatar

Member
Jan 10, 2019
2,597
lynux3 said:
I don't think anyone doubted 2GHz clock speed after a while; the doubt comes from 36CU. With PS5 seemingly being RDNA2 (no 40CU limit), having RT, allegedly supporting VRS and Oberon being associated with Navi 10 Lite by both _rogame and Komachi is where github falls apart. The only reason we think Oberon/Ariel/Flute is PS5 in the first place is because Komachi guessed it was PS5 (not being sure themselves) the same way Komachi guessed that Oberon/Ariel/Flute are Navi 10 Lite.

Needless to say, with PS5 practically being confirmed to be RDNA2, where does the Navi 10 Lite APU come into play? RDNA2 is Navi 2x.
Click to expand...
Click to shrink...

Exactly.

I also love how team GitHub just throw out any info, even from their own precious sources, that shed any doubt on a 9.2TF PS5.

Lots of info Oberon is Navi 10, from people like Komachi? IGNORE

After months and months of assurances Sony couldn't possibly develop more than one APU we now know they have a full Navi 10 line, which is the exact same as Oberon in CU counts. IGNORE

Oberon and Arden being at very different stages of development despite the belief they are both using RDNA2 and releasing at the same time. IGNORE

Insiders who have been proven to be right time and time again. IGNORE

Our own Matt telling people not to draw too many conclusions from GitHub when people were frothing about a 9.2TF PS5. IGNORE

I'm sorry but if you need to ignore so much information and so many inconsistencies that should tell you something.

There is zero evidence that Oberon is RDNA2, and far more evidence that it is an evolution of the Ariel/flute/Gonzalo line. But some people, for whatever reason, are very invested in a 9.2TF 36CU PS5.

It will be yet another loss to add to the pile for team Debbie Downer.

If PS5 is <10TF Transistor can give me an avatar of shame. I will gladly wear it.
 
ekim

ekim

Member
Oct 26, 2017
3,405
I'm really curious, what Sony's messaging will be at the reveal and if they will even talk numbers at all.
If the PS5 is more powerful than XBSX, they will probably say so. If both machines are a wash, they will surely talk about the unique features.
I remember the Xbox One reveal where Microsoft only talked about the number of transistors because they knew, XBO was less powerful than PS4 - seeing them coming out with the TFLOP number this early shows confidence.
My theory is, that Sony did not delay the reveal but was rather waiting for MS to show off first to adjust the message they want to convey.
I can also imagine they waited for MS, so they can go for a higher clock speed and better cooling to have the power advantage.
 
Sidewinder

Sidewinder

Member
Oct 25, 2017
7,189
Shambala said:
I can wait to call out certain people once the full specs are revealed. There sure are some know it all posters on this website 😂
Click to expand...
Click to shrink...

Me too, but because I've got no clue about these things I'll just point an imaginary finger at them and laugh by myself in front of my screen ;)

I just hope that PS5 will be as close as possible to Series X.
 
tzare

tzare

Banned
Oct 27, 2017
4,145
Catalunya
I was one of the ones believing both consoles would have rdna or rdna 2 arquitecture but with GCN TF in mind, for both consoles. Great to be proven wrong, but i am seriously concerned about pricing. I wonder how people believe these could can be 499 with such powerful and modern gpus on top of ssd and other improvements.
Hopefully i am wrong again
 
Axel Stone

Axel Stone

Member
Jan 10, 2020
2,771
Interesting stuff in here, especially the sums from AegonSnake, thank you for that.

It does sound like Microsoft are setting themselves up to announce a pretty chunky upclock before release, that seems to be the only explanation for that tower design from what I've read.
 
Silencerx98

Silencerx98

Banned
Oct 25, 2017
1,289
GUNDAM STARDUST said:
You guys have to end this persecution complex after months of "Phil never said 12TF Navi" talk.
Click to expand...
Click to shrink...
What makes you think it has to do with persecution complex or thinking PlayStation has it unfair? I'm just making a prediction on how this thread will go based on historical trends. I couldn't give a shit either way. As I said before, I'm an aspiring developer who hopes to work on both devices in the future. I have no horse in this race
 
klik

klik

Banned
Apr 4, 2018
873
We came a long way. A year ago most of the people thought we're gonna get GCN 8-10 teraflops
 
PJV3

PJV3

Member
Oct 25, 2017
25,676
London
Sony,Playstation, Mark Cerny and Jim Ryan, you let me down yesterday.
I might just be getting a ZX Spectrum Next instead.
 

Deleted member 12635

User requested account closure
Banned
Oct 27, 2017
6,198
Germany
Doctor Avatar said:
Exactly.

I also love how team GitHub just throw out any info, even from their own precious sources, that shed any doubt on a 9.2TF PS5.

Lots of info Oberon is Navi 10, from people like Komachi? IGNORE

After months and months of assurances Sony couldn't possibly develop more than one APU we now know they have a full Navi 10 line, which is the exact same as Oberon in CU counts. IGNORE

Oberon and Arden being at very different stages of development despite the belief they are both using RDNA2 and releasing at the same time. IGNORE

Insiders who have been proven to be right time and time again. IGNORE

Our own Matt telling people not to draw too many conclusions from GitHub when people were frothing about a 9.2TF PS5. IGNORE

I'm sorry but if you need to ignore so much information and so many inconsistencies that should tell you something.

There is zero evidence that Oberon is RDNA2, and far more evidence that it is an evolution of the Ariel/flute/Gonzalo line. But some people, for whatever reason, are very invested in a 9.2TF 36CU PS5.

It will be yet another loss to add to the pile for team Debbie Downer.

If PS5 is <10TF Transistor can give me an avatar of shame. I will gladly wear it.
Click to expand...
Click to shrink...
Good morning.

I understand you don't think Github is adequate, it can't be 36 CUs, so let's explore this.

I ask again, in good faith, what would be the config of the chip you expect?

Would you accept the BC modes to be still correct (800, 911, 2000 Mhz) - specially when looking at the HW BC Sony has a patent for?

Or do you think that the PS5 will have an Hardware Abstraction Layer (HAL) just like Xbox has which allows to do most of the BC stuff in software?

Depending on your answer and HW BC is still a thing for you, how do you think the chip is organized in SEs, SAs and CUs?

From the PS4 Pro and Github we know that for BC it has to be a multiple of 18 up to 36. So what is the number other than 36 CUs that you expect the PS5 chip has to be? Is it 54 active CUs organized in 3 SEs and 6 SAs?

I ask all this because it is really easy to say: Your (in this case my) prediction stinks without giving any information about the alternative. And with alternative I do not mean just a single number.

So what is the alternative, the not leaked chip, for PS5 in your opinion?
 
DixieDean82

DixieDean82

Banned
Oct 27, 2017
11,837
Even if it's a tiny disadvantage I still think it would be embarrassing for Xbox to get beat on power again.
 

Deleted member 27551

User requested account closure
Banned
Oct 30, 2017
660
Doctor Avatar said:
Exactly.

I also love how team GitHub just throw out any info, even from their own precious sources, that shed any doubt on a 9.2TF PS5.

Lots of info Oberon is Navi 10, from people like Komachi? IGNORE

After months and months of assurances Sony couldn't possibly develop more than one APU we now know they have a full Navi 10 line, which is the exact same as Oberon in CU counts. IGNORE

Oberon and Arden being at very different stages of development despite the belief they are both using RDNA2 and releasing at the same time. IGNORE

Insiders who have been proven to be right time and time again. IGNORE

Our own Matt telling people not to draw too many conclusions from GitHub when people were frothing about a 9.2TF PS5. IGNORE

I'm sorry but if you need to ignore so much information and so many inconsistencies that should tell you something.

There is zero evidence that Oberon is RDNA2, and far more evidence that it is an evolution of the Ariel/flute/Gonzalo line. But some people, for whatever reason, are very invested in a 9.2TF 36CU PS5.

It will be yet another loss to add to the pile for team Debbie Downer.

If PS5 is <10TF Transistor can give me an avatar of shame. I will gladly wear it.
Click to expand...
Click to shrink...
I'd take that bet to but if it is only 9.2tf give me a reason to buy the series x.
 
Raide

Raide

Banned
Oct 31, 2017
16,596
Axel Stone said:
Interesting stuff in here, especially the sums from AegonSnake, thank you for that.

It does sound like Microsoft are setting themselves up to announce a pretty chunky upclock before release, that seems to be the only explanation for that tower design from what I've read.
Click to expand...
Click to shrink...
I am guessing they will be tweaking for a while but there is only so much you can do.

Kinda interested to know which unit Phil has at home.
 
Muppet of a Man

Muppet of a Man

Member
Oct 27, 2017
7,676
My guess (again): 54 CUs (3 SEs with 18 active CUs each) clocked at 1.8 Ghz = 12.44 TFs.

Disable a single SE and 36 CUs are active for hardware-based bc mode.

Slightly narrower & slightly higher clocked than XSX.

If it is 36 CUs clocked at 2+ Ghz, they will sell it for $50.00 to $100.00 cheaper than the XSX.
 
sleepr

sleepr

Banned for misusing pronouns feature
Banned
Oct 30, 2017
2,965
Justsomeguy

Justsomeguy

Member
Oct 27, 2017
1,712
UK
Raide said:
Kinda interested to know which unit Phil has at home.
Click to expand...
Click to shrink...
It'll be a pre release "close to final" home test unit... Take home test trials are standard for xbox launches, got to get the machines rolled out globally into homes and tested with all sorts of kit and equipment and av cabinets etc. Phil just probably gets the first one but it'll doubtless be being rolled out to a limited global internal Microsoft audience over the coming months (possibly with further revisions eg alpha unit, beta unit, final retail unit etc)
 
Doctor Avatar

Doctor Avatar

Member
Jan 10, 2019
2,597
Colbert said:
Good morning.

I understand you don't think Github is adequate, it can't be 36 CUs, so let's explore this.

I ask again, in good faith, what would be the config of the chip you expect?

Would you accept the BC modes to be still correct (800, 911, 2000 Mhz) - specially when looking at the HW BC Sony has a patent for?

Or do you think that the PS5 will have an Hardware Abstraction Layer (HAL) just like Xbox has which allows to do most of the BC stuff in software?

Depending on your answer and HW BC is still a thing for you, how do you think the chip is organized in SEs, SAs and CUs?

From the PS4 Pro and Github we know that for BC it has to be a multiple of 18 up to 36. So what is the number other than 36 CUs that you expect the PS5 chip has to be? Is it 54 active CUs organized in 3 SEs and 6 SAs?

I ask all this because it is really easy to say: Your (in this case my) prediction stinks without giving any information about the alternative. And with alternative I do not mean just a single number.

So what is the alternative, the not leaked chip, for PS5 in your opinion?
Click to expand...
Click to shrink...

You're perfectly entitled to your opinion! But at least acknowledge there is plenty of possibility anyone claiming a 36CU 2.0Ghz will turn out to very wrong. Indeed as time goes it seems the potential for that set up to be wrong increases.

Lots of Navi 10 chips from Sony, so once we know that multiple different archs and APUs are being developed anything is possible. There is also the question of why they bothered developing RDNA1 parts.

My prediction is 3SEs, clover design as the natural evolution of the PS4Pro's butterfly design for easy BC but also not crippling console performance for it.
 
Pryme

Pryme

Member
Aug 23, 2018
8,164
N.47H.4N said:
lol some guys just can't accept PS5 will be one par next gen with XsX and want to believe in that 3TF deficit.
Click to expand...
Click to shrink...

we aren't at the point to make definitive 'will' statements, ya know.
Good to have the info that they're the same architecture though. Helps to discard the Github stuff and should narrow whatever gap there is between the performance of both consoles.
 
EvilBoris

EvilBoris

Prophet of Truth - HDTVtest
Verified
Oct 29, 2017
16,684
Doctor Avatar said:
My prediction is 3SEs, clover design as the natural evolution of the PS4Pro's butterfly design for easy BC but also not crippling console performance for it.
Click to expand...
Click to shrink...
That would make sense if it was a machine trying to to be 3Xthat of a PS4.
Otherwise , I'm not sure why a "clover" design would be relevant as a next gen machine.
 
chris 1515

chris 1515

Member
Oct 27, 2017
7,074
Barcelona Spain
Again Oberon is Navi 10 see PCI ID 13F9 , here is the proof it can't be RDNA 2.

forum.beyond3d.com

Baseless Next Generation Rumors with no Technical Merits [post E3 2019, pre GDC 2020] [XBSX, PS5]

I don't see how any third party dev would be privy to any changes to the hardware regarding cooling. Yes they will obviously have performance targets and so forth but I get very sceptical when they start talking about design decision that only the hardware guys at AMD and Sony would know. Even...
forum.beyond3d.com forum.beyond3d.com
 

Deleted member 12635

User requested account closure
Banned
Oct 27, 2017
6,198
Germany
Doctor Avatar said:
You're perfectly entitled to your opinion! But at least acknowledge there is plenty of possibility anyone claiming a 36CU 2.0Ghz will turn out to very wrong. Indeed as time goes it seems the potential for that set up to be wrong increases.

Lots of Navi 10 chips from Sony, so once we know that multiple different archs and APUs are being developed anything is possible. There is also the question of why they bothered developing RDNA1 parts.

My prediction is 3SEs, clover design as the natural evolution of the PS4Pro's butterfly design for easy BC but also not crippling console performance for it.
Click to expand...
Click to shrink...
Thanks for the response though what is the native clock instead? And why so angry?

Also, you must have missed the part where other members posted that Navi 10 was related to Ariel/Gonzalo and not Flute/Oberon. Posted by anexanhume and DrKeo.

Bookmarked
 

Dekim

Member
Oct 28, 2017
4,300
I never understood the logic behind the theory that Sony either created their own RT solution or was using a third party's. Just think about it.

Why would Sony add cost and complexity to their console by either investing in their own RT tech or using some third party's RT tech? Especially for a feature Sony isn't touting in marketing? This is in stark contrast to their SSD solution where I can easily see them using a unique, custom SSD solution of their own making because of how much they emphasized how important the PS5's SSD will be. It is even sillier to think about when Sony's main vendor--AMD--have their own RT tech baked into the technology architecture Sony will use from them.

The notion that Sony would use a third party's RT tech or invent their own never made sense.
 
EvilBoris

EvilBoris

Prophet of Truth - HDTVtest
Verified
Oct 29, 2017
16,684
NoUse4AName said:
Do we have any info about how PS5's BC will work....or is just more speculation?
Click to expand...
Click to shrink...

There was one leak that showed that the hardware could had 2 modes lock itself to the same specifications as the PS4 and the PS4 pro.
This suggests they are going for a faithful reproduction of what a standard PS4/Pro does rather than going further and trying to improve IQ or framerate further.
 
chris 1515

chris 1515

Member
Oct 27, 2017
7,074
Barcelona Spain
Colbert said:
Thanks for the response though what is the native clock instead? And why so angry?

Also, you must have missed the part where other members posted that Navi 10 was related to Ariel/Gonzalo and not Flute/Oberon. Posted by anexanhume and DrKeo.

Bookmarked
Click to expand...
Click to shrink...

lol

forum.beyond3d.com

Baseless Next Generation Rumors with no Technical Merits [post E3 2019, pre GDC 2020] [XBSX, PS5]

I don't see how any third party dev would be privy to any changes to the hardware regarding cooling. Yes they will obviously have performance targets and so forth but I get very sceptical when they start talking about design decision that only the hardware guys at AMD and Sony would know. Even...
forum.beyond3d.com forum.beyond3d.com

Here Oberon is link to Flute and a PCI ID use for Navi10 Lite.

www.guru3d.com

ZEN2 8c/16t with NAVI GPU Spotted - AMD Console chip?

Twitter user Komachi has stumbled into a quite remarkable finding, actually, he might have stumbled into the pending Xbox (XBox Scarlett) APU. According to the leaked info, this is a Zen 2 architect...
www.guru3d.com www.guru3d.com

EDIT:


EMyJPSPUcAA-W-L.png


EDIT:

EMyTgjpW4AAw-4F


This is impossible Oberon is RDNA 2 without RT & VRS. What is a RDNA 2 GPU without RT and VRS maybe a RDNA1 GPU.
 
Last edited:
Axel Stone

Axel Stone

Member
Jan 10, 2020
2,771
Raide said:
I am guessing they will be tweaking for a while but there is only so much you can do.
Click to expand...
Click to shrink...

I'm not so sure this time around based on AegonSnake's sums.

If the power draw of 12 RDNA2 TF's really is equivalent to modern consoles, then designing your box like a wind tunnel is overkill.

Also, at 54-56 CUs, the current best guesses for die size, I think that the estimates for clock speed were around 1700mhz, but it looks like RDNA2 can be comfortably clocked higher. Doing that would decrease yields, sure, but Microsoft do seem to have their sights set on the top end of the market with XSX.

The design almost seems to be based around giving headroom for an upclock
 
Thera

Thera

Banned
Feb 28, 2019
12,876
France
Doctor Avatar said:
There is zero evidence that Oberon is RDNA2, and far more evidence that it is an evolution of the Ariel/flute/Gonzalo line. But some people, for whatever reason, are very invested in a 9.2TF 36CU PS5.
Click to expand...
Click to shrink...
Yep.
If you believe Jason (and who wouldn't), you know it won't be 9.2 since the moment it popped.
So I don't think you should try to convince other about GithHub who already have their mind set.
Doctor Avatar said:
You're perfectly entitled to your opinion! But at least acknowledge there is plenty of possibility anyone claiming a 36CU 2.0Ghz will turn out to very wrong. Indeed as time goes it seems the potential for that set up to be wrong increases.

Lots of Navi 10 chips from Sony, so once we know that multiple different archs and APUs are being developed anything is possible. There is also the question of why they bothered developing RDNA1 parts.

My prediction is 3SEs, clover design as the natural evolution of the PS4Pro's butterfly design for easy BC but also not crippling console performance for it.
Click to expand...
Click to shrink...
Like I said, don't try to convince them, it won't work and will exhausted you.
 
Sekiro

Sekiro

Member
Jan 25, 2019
2,938
United Kingdom
So is BC compatibility and cost saving the two main factors as to why Sony is going with 36CU if the github leak is true?

Can't think of any other reason they would go with that CU number.
 
chris 1515

chris 1515

Member
Oct 27, 2017
7,074
Barcelona Spain
Thera said:
Yep.
If you believe Jason (and who wouldn't), you know it won't be 9.2 since the moment it popped.
So I don't think you should try to convince other about GithHub who already have their mind set.

Like I said, don't try to convince them, it won't work and will exhausted you.
Click to expand...
Click to shrink...

We can tell to everyone with a 5700 and 5700 XT you have a RDNA 2 GPU. Oberon has no RT and VRS this is not a RDNA2 GPU.
 
Reader mode
Status
Not open for further replies.