• 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.

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.

LordBlodgett

Member
Jan 10, 2020
806
XSX is roughly equivelant to 17.25 TF GCN

and for lockhart, 4 TF would be 5.75 TF (so lockhart, if it exists, would probably be 4.2TF or higher as it doesn't fully line up with X1X)

PS5 @ 9.2 TF would be 13.225 TF

also MykhellMikado sorry you were right they did announce 15% perf improvement over RDNA1
Wait, when did they announce 15% IPC improvement over RDNA1? I watched the whole presentation and the QA and I didn't see that.....
 

Proven

Banned
Oct 29, 2017
5,841
Fair.. with Sony's silence ofcourse people will speculate on anything and rumors spread like wildfire.

I have seen it happen on different products whenever a company is silent.

Yeah, I'm only using Github as a source because I'm more inclined to believe data mined info directly from AMD that matches a code name that aligns with the PS5 over anything else. There really isn't anything else to speculate off of.
 

ILikeFeet

DF Deet Master
Banned
Oct 25, 2017
61,987
Wait, when did they announce 15% IPC improvement over RDNA1? I watched the whole presentation and the QA and I didn't see that.....
they didn't. people are assuming it's the same 15% (actually 14%) performance improvement that RDNA1 had over GCN because that had the same 50% perf/watt improvement
 

Vimto

Member
Oct 29, 2017
3,714
Yeah, I'm only using Github as a source because I'm more inclined to believe data mined info directly from AMD that matches a code name that aligns with the PS5 over anything else. There really isn't anything else to speculate off of.

are you participating in the "klee vs github" bet by any chance? :P
 

Deleted member 19767

User requested account closure
Banned
Oct 27, 2017
2,098
Imagine waking up every morning and saying to yourself, 'I'm going to bring up Github again today'.
What a world we live in.
 

DopeyFish

Member
Oct 25, 2017
10,783
they didn't. people are assuming it's the same 15% (actually 14%) performance improvement that RDNA1 had over GCN because that had the same 50% perf/watt improvement

no

there was a 25% IPC improvement on RDNA over GCN

someone misheard (not me) the RDNA2 IPC improvement over RDNA1 and i accidentally relayed that misheard conversation and have since corrected.
 

DrScissorsMD

Attempted to circumvent ban with alt account
Banned
Jan 19, 2019
564
Github Was PS4 Pro BC testing.

There really isn't anything more to say here as it's as obvious as the sun is bright, especially after the AMD show.
Is that not weird though, to have PS4Pro testing in with next gen tests? Or are you saying they are testing for a PS4Pro mode for PS5? Not saying you're wrong at all as I don't have the technical know-how to say either way.
 
Oct 25, 2017
9,205
9.2 TF beats Stadia's 10.7, and aiming for does not always mean it succeeded.

The biggest problem right now is that the insiders that have spoken are either questionable sources or are not exactly clear with what they're saying. So that's why there's so much back and forth.

Literally nothing is questionable about what Matt said. He said the difference in power is insignificant. So did KLee. ZhugEx said the BOM are close. But keep listening to and quoting the dregs of Twitter instead.
 

III-V

Member
Oct 25, 2017
18,827
One thing I find really interesting in the GitHub docs is that they ARE STILL WORKING ON ARIEL_B0...... but why would they do that if we already have Oberon_B0. Ariel and Oberon are both fairies... Dual GPU theory rises from the grave?

r197471 | name withheld | 2019-05-16 13:03:31 +0800 (Thu, 16 May 2019) | 2 lines

fix ariel b0 build error: array subscript is above array bounds [-Werror=array-bounds]

r197498 | name withheld | 2019-05-16 23:27:17 +0800 (Thu, 16 May 2019) | 1 line

[Peak3D] Update CRCs for Oberon
 

Electro

Banned
Oct 25, 2017
5,899
Vienna
User banned (permanent): platform wars, antagonising other uses over multiple posts, long history of bans for platform wars
After a look who these GitHub guys are, I saw they are owned since 2018 by MS.

Why should I believe them more than trustable insiders ? xD
 

jroc74

Member
Oct 27, 2017
28,992
How would this be for a hype moment...

PS5 reveal (PlayStation Meeting, pre-recorded, not live). Cerny presentation, developers talk, controller and console revealed.

11.4 TF.

Demon's Souls Remake. Insomniac new Ratchet & Clank that looks almost as good as a Pixar film

FFVIIR BC enhancements (native 4K, 60fps)

A few clips of Gran Turismo 7 with ray tracing.

Monster Hunter World and Iceborne BC enhancements

and finally, Capcom Presents

D E E P - D O W N - showing incredible flame physics and gameplay that visually far surpasses the 2013 trailer that ran on a 3TF GTX 680.
Hype thru the roof for me.
 

DopeyFish

Member
Oct 25, 2017
10,783
One thing I find really interesting in the GitHub docs is that they ARE STILL WORKING ON ARIEL_B0...... but why would they do that if we already have Oberon_B0. Ariel and Oberon are both fairies... Dual GPU theory rises from the grave?

not that i'm paying much attention to this

isn't ariel part of oberon?

After a look who these GitHub guys are, I saw they are owned since 2018 by MS.

Why should I believe them more than trustable insiders ? xD

Github is a repository service, MS puts stuff on their under their own repositories

it was an AMD employee who was using it and someone found it and AMD forced it to be taken down

this is much like someone putting something on linkedin or whatever. just because MS owns it means nothing.
 

Proven

Banned
Oct 29, 2017
5,841
are you participating in the "klee vs github" bet by any chance? :P

No, because there's too much info in the github data. The PS5 could end up being 36 CUs and end up being above 9.2 TF. I think the PS5 ends up hitting 10 TF if they really want to hit that number.

I am also 50/50 that github even means anything but there's really nothing else to speculate off of.
 

Deleted member 135

User requested account closure
Banned
Oct 25, 2017
11,682
Hmm okay so I assume you're partly basing that off there being no VRR or ray-tracing, which there wouldn't be if you were running back compat games? That would make sense.
Yep. I imagine that there will also be a PS5 Boost Mode that enables brute forcing even better performance from PS4 games like the PS4 Pro Boost Mode does.

What is interesting is what kind of extra features they could overlay on PS4 games with a boost mode.
 

LordBlodgett

Member
Jan 10, 2020
806
Isn't Series X also RDNA2 though....?
Yes, but it is a big boy chip at 400mm^2. Even with lower clock speeds it's still going to a big cooler. The other part about that is cost and noise. Larger fans are usually quieter, and the tower design means they don't need to do anything special like the vapor chamber they cooked up for Xbox One X, meaning low costs for the cooler as well.
 

DavidDesu

Banned
Oct 29, 2017
5,718
Glasgow, Scotland
Regarding disc-less Lockhart I think it could be the wise move. Provided it has a 1TB SSD of course.

For my use case I'm getting a PS5 no matter what. If I can also get a next gen Xbox for cheap to play Gamepass titles only (the "All Inclusive" contract option which is pretty affordable) then I'm probably not actually going to be buying many or any other titles outside of the Gamepass selection. Lockhart basically is a cheaper "in" to all the Xbox exclusives, and still a great place to play whatever third party titles Gamepass will include. I won't really be worrying about it having a disc drive.

Since Microsoft seem to have a business plan which priorities Gamepass subs over all else then a "cheap" next gen Xbox as a Gamepass machine makes sense. I'm still going with Sony too but Microsoft will now get some of my money rather than none, which would be the case if I had to choose one expensive next gen console completely over the other (and I'd choose PS5).

I've really not cared about Xbox this generation at all. Blame the initial bad PR blunders but then also the lack of really quality titles (in my eyes) versus PS4. However I'm really starting to change. Seeing that I can get an Xbox One X and Gamepass for two years for £25ish a month, and knowing I'll want to play some Halo and Forza next gen, I'm looking forward to being in both the main ecosystems rather than choosing just one this time around. I laughed at Lockhart initially, but if they can make it affordable then frankly, I'm in. Oh and Flight Simulator is literally the key here. I want to play that game so badly.
 

jroc74

Member
Oct 27, 2017
28,992
9.2 TF beats Stadia's 10.7, and aiming for does not always mean it succeeded.

The biggest problem right now is that the insiders that have spoken are either questionable sources or are not exactly clear with what they're saying. So that's why there's so much back and forth.
And the github data isn't questionable?

I would think after yesterday it would be even more but yet....
 

III-V

Member
Oct 25, 2017
18,827
not that i'm paying much attention to this

isn't ariel part of oberon?
nope.

Ariel is the iGPU of Gonzalo APU.
Oberon is the iGPU of Flute APU


In other news: Proof that Oberon at time of GitHub testing is NAVI10 i.e. RDNA 1.0? (gfx10_ libs) ?
r197717 | name withheld | 2019-05-20 22:29:45 +0800 (Mon, 20 May 2019) | 2 lines

confirmed with SIE, remove external/amd/lib/gfx10_a and external/amd/lib/gfx10_b/cmodel
 

Brees2Thomas

Member
Dec 27, 2019
1,525
Imagine waking up every morning and saying to yourself, 'I'm going to bring up Github again today'.
What a world we live in.
The mystery over the GitHub leak is the most interesting thing by far in these speculation threads IMO. I can see both sides of the argument and there are valid points being made by both camps. I am fascinated to learn the truth about how it relates to the PS5. I couldn't imagine what we'd be talking about all this time with no news to speak of. Certainly we wouldn't be close to OT 12 by now, I don't think.
 

Lorul2

Member
Jan 4, 2018
768
Wait so where are we now with these consoles I'm confused? who has what terra-flops and why?
 

PJV3

Member
Oct 25, 2017
25,676
London
The mystery over the GitHub leak is the most interesting thing by far in these speculation threads IMO. I can see both sides of the argument and there are valid points being made by both camps. I am fascinated to learn the truth about how it relates to the PS5. I couldn't imagine what we'd be talking about all this time with no news to speak of. Certainly we wouldn't be close to OT 12 by now, I don't think.

I don't mind it, but I find both sides pretending they don't know the answers to their questions that they have asked a hundred times by now amusing.
 

LordBlodgett

Member
Jan 10, 2020
806
I do have a question now...

Will Lockhart have rt and vrs? It's supposed to be aiming for 1440p. There are gpus out that do that, on RDNA 1 tho.

Or are we looking at an RDNA 1 with RDNA 2 features?
I would guess that they would also develop Lockhart on the same architecture as its big brother, the Series X. No reason to think they would go RDNA1. There wouldn't be much in the way of cost savings unless we are talking about different process nodes (7nm vs 7nm+ for example.)
 

III-V

Member
Oct 25, 2017
18,827
May I ask why?
I mean, it's not like the die space where you would expect a fourth SE to be located could not be used for anything else in these custom APUs, huh?
XSX gets to 56+ CU's with 2 SE's - no need to add a SE.

GitHb has test results going all the way down to an individual WGP (dual CU) - no need to have multiples of 18 past 36 CU.
 

LordBlodgett

Member
Jan 10, 2020
806
Sony are silent because they already know MS are beat. They will drop a 13.2tf console at $450, they have the better games, they have a huge install base, the next gen starts when they say so, they have the entire industry on strings.

If they were lower in power than Microsoft they would have been dropping news left right and centre to mitigate it and to try and cover the 12 tf news about XSX.

Lockheart will be released by MS to try and play catch up on Sony, and will fail again.
Beautiful Console Warring! A masterpiece, I must say!
 

Puroresu_kid

Banned
Oct 28, 2017
9,465
Sony are silent because they already know MS are beat. They will drop a 13.2tf console at $450, they have the better games, they have a huge install base, the next gen starts when they say so, they have the entire industry on strings.

If they were lower in power than Microsoft they would have been dropping news left right and centre to mitigate it and to try and cover the 12 tf news about XSX.

Lockheart will be released by MS to try and play catch up on Sony, and will fail again.

I know I'm late but this post is embarrassing.
 

Ateron

Member
Oct 30, 2017
32
For MS, I think a 9.2 TF machine at $400 is much scarier competition than a 12-13 TF Sony console at $500 (which is where I presume the XSX will be).

I don't want to be scared, I want to be wowed. Let them do the number crunches, it has no bearing on us, the consumers. Either way Sony will be fine, just want both to be specced similarly, doesn't matter who "wins" as long as they are perfectly comparable. 9.2 wouldn't be the end of the world but 12 is more tempting for me
 

LordBlodgett

Member
Jan 10, 2020
806
One of the best break downs I've seen posted yet, thanks for this.

I would have also added that people kept saying "clover configuration" to refer to the possibility of Sony using 3 or 4 SE's with 18 CU's each to = 54 or 72 (lol) total, and that the supposed Github leak was only showing 36 because 1 or 2 SE's were disabled at the time of testing.

Sorry to muddy the waters but just though it was an interesting tidbit/theory that expands on the "PS5 has to use multiples of 18" concept to help people understand why they keep seeing the 18/36/54/72 numbers when we speculate.
Yeah, I didn't include any of the "butterfly" or "clover" configuration stuff......Just trying to keep it somewhat simple. Still a long post :)
 

jroc74

Member
Oct 27, 2017
28,992
nope.

Ariel is the iGPU of Gonzalo APU.
Oberon is the iGPU of Flute APU


In other news: Proof that Oberon at time of GitHub testing is NAVI10 i.e. RDNA 1.0? (gfx10_ libs) ?
So....where is that RDNA 2 PS5 gpu at......

Hey, RDNA 1 plus RDNA 1... two gpus....

Does my math work out?
 

Doctor Avatar

Member
Jan 10, 2019
2,591
Klee literally quoted someone who said 12 TF right after Windows Central got the scoop about the boxes specs months before Xbox ever confirmed. I'm not sure how repeating what someone else already said is confirming it.

And I'm honestly just wondering if he meant Stadia's 10.7 number in terms of GCN or RDNA2. If he means RDNA2 then I don't think Github is as accurate.

Klee confirmed it and said it wasn't surprising and changed nothing before MS did. It fell completely in line with his predictions too.

I think a lot of people here are going to owe him an apology.
 

III-V

Member
Oct 25, 2017
18,827

Oberon and Ariel are both iGPU. Ariel was/is the progenitor to Oberon. Oberon is being regression tested to Ariel.
Ariel is the folder root for nearly everything NAVI10 related, as far as I can see. Ariel is very old.

Komachi found the GitHub leak as this is literally word-for-word in the interns "daily_work" task log, as instruction from a superior.

For Oberon:
git fetch ssh://---withheld----/diagnostics/ec/diag_gpu_ariel refs/changes/---withheld
 

dom

▲ Legend ▲
Avenger
Oct 25, 2017
10,438
Yeah, I'm only using Github as a source because I'm more inclined to believe data mined info directly from AMD that matches a code name that aligns with the PS5 over anything else. There really isn't anything else to speculate off of.
If there is nothing to speculate off of, why are you constantly hammering github and then go on to post timdog and friends twitter posts?
 

Sekiro

Member
Jan 25, 2019
2,938
United Kingdom
PS4 Pro has a 2.13ghz CPU and a 36CU 911mhz GPU.

When playing unpatched base PS4 games with Boost Mode disabled the CPU runs at 1.6ghz and the GPU only actives 18CUs and runs them at 800mhz to achieve practically perfect compatibility.

When Boost Mode is enabled the CPU runs at the full 2.13ghz speed and the GPU runs at the full 911 mhz speed, but the number of active CUs remains at 18.

When a patched game is played the CPU runs at the full 2.13ghz and all 36 CUs are active and run at 911mhz.


The only logical conclusion to make is, and I dare say Occam's Razor says, that any information about the PlayStation 5 utilizing any of the aforementioned numbers, whether clock speeds or CU counts, is only and purely indicative of compatibility modes for PS4 and PS4 Pro games.
I could see that happening:
PS5 GPU:
18CU = PS4 Mode
36CU = PS4 Pro mode
56CU = Native PS5 games
 
Status
Not open for further replies.