• 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 & Next Xbox |OT| Speculation/Analysis/Leaks Thread

Next Gen Consoles Power Level


  • Total voters
    237

MadAnon

Member
Sep 11, 2018
186
122
230
If I understand correctly those gen0, gen1 are not just reference to ps4 and pro clocks but actually clock speed modes of this GPU, with gen2 being PS5 mode? Holy shit!!??
 
Last edited:
  • Like
Reactions: joshwaan

Sycomunkee

Member
Jan 21, 2019
1,459
1,328
370
The power of what?
I don't like reporting people for "trolling", but what the FUCK is this shit?
Did you read the post I was quoting? I didn’t make it up. Report away, I’d be Concerned for everyones sake if i was banned for something so minimal such as directly responding to somebody’s post. Let alone posting my opinion on a forum.
 
Last edited:

DeepEnigma

Gold Member
Dec 3, 2013
23,382
22,498
1,045
If I understand correctly those gen0, gen1 are not just reference to ps4 and pro clocks but actually clock speed modes of this GPU? Holy shit!!??
That would make sense for BC, since they are the clocks of the each "gen" on the platform/ecosystem.

Gen 0 profile is PS4 clock.
Gen 1 profile is Pro clock.
Gen 2 profile is PS5 clock.
 
Last edited:

sonomamashine

Member
Jun 29, 2019
503
240
265
I thought someone posted a few weeks back (or it could have been a few months back at this point), that it's a 1,40 IPC boost. Or was the 1.40 IPC boost the max and the 1.25 IPC boost the average?
maybe they got it mixed, 1.5x increase in performance per watt over existing Vega GPUs, Navi is 25% faster in terms of IPC and 50% more efficient.
 

SonGoku

Gold Member
Aug 16, 2018
3,849
3,651
560
11TF in the new RDNA counting numbers? Isn't that equivalent to 15 TFs if you couldn't the old\current way?
Pointless to compare to GCN, VEGA hit big diminishing returns
RDNA is now on par with Turing perf/fp, that's a better point of comparison

11TF RDNA = ~2080S
 
Last edited:

Thedtrain

Member
Jul 30, 2018
52
95
195
I’m dying for R600’s return.

On topic: So if this is Sony chip, I feel like I’m going to be let down. Going small on the CU count and ramping up the clocks.

9.2tf depresses me
 
  • Triggered
Reactions: joshwaan

TLZ

Member
Oct 20, 2015
6,265
3,568
740
Sydney
I would think at least more than the Pro, no?
Ok, so here's a question for the more knowledgeable dudes and dudettes in here. Since the PS4 has 18 CUs, and the Pro has 36 CUs, double the PS4 to maintain BC, wouldn't the PS5 continue the trend and have either 3/4x the PS4's CUs to maintain full BC?

So either 18x3=54 CUs or 18x4=72 CUs? And at that 2ghz clock, either 13.8tf (this one probably looks safer) or 18.4tf (this one looks too high)?
 
Last edited:

vpance

Member
Jun 20, 2005
7,094
318
1,380
If they went 2ghz, a high CU count is improbable even if on N7P or EUV. We know RDNA skyrockets in watts beyond 1.8ghz.

Sony trollin Komachi with these WIP one off GPUs IMO.
 

DeepEnigma

Gold Member
Dec 3, 2013
23,382
22,498
1,045
Ok, so here's a question for the more knowledgeable dudes and dudettes in here. Since the PS4 has 18 CUs, and the Pro has 36 CUs, double the PS4 to maintain BC, wouldn't the PS5 continue the trend and have either 3/4x the PS4's CUs to maintain full BC?

So either 18x3=54 CUs or 18x4=72 CUs? And at that 2ghz clock, either 13.8tf (this one probably looks safer) or 18.4tf (this one looks too high)?
What you are saying could have some merit. I don’t think it will be 2K clocks, 1800 max. Probably more in the 1600 range.
 
  • Thoughtful
Reactions: TLZ

DemonCleaner

Member
Jun 11, 2004
1,653
279
1,625

@7nm EUV the power/voltage frequency curve would probably shift to the right. At least is you wouldn't increase density to the max.

tbh I'm coming round on this pastebin https://pastebin.com/y8qXme7b . ignoring the devkit and looking at the retail specs it looks like it could happen.

I think some of the text in there is marketing spin. but 275W PSU excites me
pastebin leak is a lie. looks plausible at first, but there is no way it would just have 560GB/s of memory bandwidth at 384bit bus (which it necessarly need to have with 12x 2GB GDDR6 modules).




Too pessimistic - real world switched mode supplies easily get over 90% and 95% isn't unrealistic .. that'l get you under the magic 200W figure.. :)
That's awesome to hear. Why are 80plus gold PSUs for the PC still that expensive then?

Looking at your results an APU die with 60CU iGPU @1600 would comfortably hit 150W or less on N7P
would maybe also be (boarderline) possible even on normal 7nm at a bit lower clocks :messenger_horns:
 
Last edited:

Mass Shift

Member
Jun 12, 2019
176
126
285
I meant to say consoles won't sport less than 48CUs
2GHz is unrealistic for consoles be it 36CU or 48CU
It doesn't sound unrealistic if as a result of the 7nm EUV process, the power efficiency makes allowances for the 2GHz clock.

I don't know SonGoku, I pray that it's at least 48. But Leadbetter, DF, Gonzalo, it's beginning to add up with a fair bit of consistency.
 

xool

Member
May 29, 2018
841
723
335
That is retarded. Different nodes have different voltage/power curves and different architectures have different frequency/voltage curves.
No it is not "retarded" because I was comparing performance vs clock - power usage didn't even come into it.

Seriously don't go throwing around phrase "retarded"

Now comparing outright performance is one thing as you can get perf/transistor and comparing Tflops to find perf/tflop is also useful but using the full fat die of one generation Vs the cut down version of another and saying perf/transistor has regressed when you don't even know how many transistors have been turned off is fucking moronic.
1080 is full fat, 5700 is 36 of a max 40 CU.

Again lay off the attacks.
 

xool

Member
May 29, 2018
841
723
335
pastebin leak is a lie. looks plausible at first, but there is no way it would just have 560GB/s of memory bandwidth at 384bit bus (which it necessarly need to have with 12x 2GB GDDR6 modules).
Samsung GDDR6 gos as low as 12Gb/s/pin - that gives nearly 560GB/s (576) on a 384bit bus .. they would be cheapest GDDR6 .. just saying - I don't really believe the numbers either.
 

xool

Member
May 29, 2018
841
723
335

I'm at the stage of "don't believe any of this". Komachi et al is/are feeding everyone a line - snippets of manufactured info that looks like Intel.
  • Downclock to PS4/Pro GPU speeds for what ? - backwards compatability? on a new sub-architecture, with a completely different CPU design - nope
 
Dec 12, 2006
2,802
374
1,080

I'm at the stage of "don't believe any of this". Komachi et al is/are feeding everyone a line - snippets of manufactured info that looks like Intel.
  • Downclock to PS4/Pro GPU speeds for what ? - backwards compatability? on a new sub-architecture, with a completely different CPU design - nope
Can't understand why people put any stock into anything that Komachi or Apisak tweets about next gen. It's like the blind leading the blind.
 
Last edited:

llien

Gold Member
Feb 1, 2017
5,724
2,822
720
...and the XT doesn't [obviously] beat the 2070, despite being higher clocked. [edit I accept it's +5% better]...
It is hilarious in how many ways you manage to be wrong in one sentence.





NVs actual clock is much higher than claimed.
 
Last edited:

xool

Member
May 29, 2018
841
723
335
It is hilarious in how many ways you manage to be wrong in one sentence.
..

NVs actual clock is much higher than claimed.
Excuse me for taking nvidia's and every other resellers specs as truth .. Oh wait you;re quoting maximum overclock - vs 5700XT standard clocks

Try linking some real figures .. like for like.
 

xool

Member
May 29, 2018
841
723
335
They have an actual track record unlike you or SonGoku. And those are not blind guesses they share.
- they don't seem to have the claimed track record at all - supposedly they knew the PS4 and PS4-Pro chip IDs - but where was this info before release .. it didn;t exist - the first time we heard about this supposed "track record" was long after both consoles where out, and the info in the public domain..

An article at Eurogamer fed into the hype https://www.eurogamer.net/articles/digitalfoundry-2019-is-amd-gonzalo-the-ps5-processor-in-theory

I'll take it all back if anyone can prove either had any info on PS4/PS4-Pro before release. Not reverse engineered claims of insight..

Bring it.
 

R600

Member
Jun 17, 2019
415
250
215
- they don't seem to have the claimed track record at all - supposedly they knew the PS4 and PS4-Pro chip IDs - but where was this info before release .. it didn;t exist - the first time we heard about this supposed "track record" was long after both consoles where out, and the info in the public domain..

An article at Eurogamer fed into the hype https://www.eurogamer.net/articles/digitalfoundry-2019-is-amd-gonzalo-the-ps5-processor-in-theory

I'll take it all back if anyone can prove either had any info on PS4/PS4-Pro before release. Not reverse engineered claims of insight..

Bring it.

I'm at the stage of "don't believe any of this". Komachi et al is/are feeding everyone a line - snippets of manufactured info that looks like Intel.
  • Downclock to PS4/Pro GPU speeds for what ? - backwards compatability? on a new sub-architecture, with a completely different CPU design - nope
Yea, apparently for BC. As you can see, they downclock Jaguar in Pro as well as PS4Pro GPU, along with deactivating half of CUs, which interestingly matches exact specs of PS4.

“To give an example, the GPU of the prior version of the system might run at a GPU clock of 500 MHz, and the current system might run at a GPU clock [156] of 750 MHz. The system would run with [156] set to 750 MHz when an application is loaded that is designed only for the current system. In this example, the cycle counter [CC] would correspond to the 750 MHz frequency (i.e., it is a true cycle counter). When a legacy application (i.e., an application designed for the prior version of the system) is loaded, the system [100] may run at a frequency slightly higher than the operating frequency of the prior system (e.g., with [156] set to 505 MHz). In this backward compatible mode, the GPU spoof clock [135] would be configured to run at 500 MHz, and the cycle counter CC would be derived from the spoof clock, thus providing the expected value to the legacy application.”

Maybe matching clocks is important even for gen to gen emulation?

Whatever he provided, from AMD CPUs to GPUs is certainly better "guess" then any of you guys have had up until now.
 

llien

Gold Member
Feb 1, 2017
5,724
2,822
720
Oh wait you;re quoting maximum overclock
Grasping straws you are (oh look, the same 1935Mhz and 1845Mhz sustained as above):










 
Last edited:

CSmath22

Neo Member
Aug 8, 2019
96
98
140
Excuse me for taking nvidia's and every other resellers specs as truth
You wouldn't know unless you've had one (and I don't mean that in an elitist way.) My nvidia GPU is listed with a 'boost clock' of 1680 MHz, but will genuinely run around 1845-1860 MHz not overclocked (and sliders left at default.)
 
Last edited:
  • Thoughtful
Reactions: SonGoku

mckmas8808

Member
May 24, 2005
41,637
4,422
1,630

I'm at the stage of "don't believe any of this". Komachi et al is/are feeding everyone a line - snippets of manufactured info that looks like Intel.
  • Downclock to PS4/Pro GPU speeds for what ? - backwards compatability? on a new sub-architecture, with a completely different CPU design - nope
How many TFs would the PS5 have if the GPU ran at an impossible 2 Ghz and it was 48 CUs?
 

LordOfChaos

Member
Mar 31, 2014
9,121
1,173
710
So it doesn't seem like they truly want to tariff things like phones and consoles, being that there's several American megacorps impacted by those


How many TFs would the PS5 have if the GPU ran at an impossible 2 Ghz and it was 48 CUs?
12.1
 
Last edited:
  • Like
Reactions: mckmas8808

mckmas8808

Member
May 24, 2005
41,637
4,422
1,630
So it doesn't seem like they truly want to tariff things like phones and consoles, being that there's several American megacorps impacted by those




12.1
LOL @ Trump. He's not serious about this tariff and I'm happy about that. I want a monster PS5!
 

xool

Member
May 29, 2018
841
723
335
Grasping straws you are (oh look, the same 1935Mhz and 1845Mhz sustained as above):










Dude I don't know what the fuck you are talking about - the comparison was with a 2070 not 2070 super .. not continuing this .. go away
 

SonGoku

Gold Member
Aug 16, 2018
3,849
3,651
560
It doesn't sound unrealistic if as a result of the 7nm EUV process, the power efficiency makes allowances for the 2GHz clock.

I don't know SonGoku, I pray that it's at least 48. But Leadbetter, DF, Gonzalo, it's beginning to add up with a fair bit of consistency.
Makes no sense to clock all the way up to 2GHZ and lowball CU count, complete opposite of console design paradigms
They'll get a more efficient design by compromising a little on clock speeds, say 1.8GHZ and go with 48CU or if that's too power hungry, 56CU - 1.6GHz

Not to mention EUV density increase enables a 64 to 72CU part.
Suit yourself but I'm fallowing komachi leaks. Things constantly add up but you choose to ignore it.
Its the other way around actually: mental gymnastics, speculation and constant mix and matching
I treat it for what is: rumor and speculation.
 
Last edited:
  • Like
Reactions: sonomamashine