Going and 2Ghz would also be requiring a big investment in cooling .
I mean just look at AMD current cards and they not 2Ghz .
Also yields at that speed can also become a problem .
-
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.
Next-gen PS5 and next Xbox speculation launch thread |OT8| - The Dark Tower (See Staff Post)
- Thread starter Mecha Meister
- Start date
- Status
Going and 2Ghz would also be requiring a big investment in cooling .
I mean just look at AMD current cards and they not 2Ghz .
Also yields at that speed can also become a problem .
You're forgetting that sony helped AMD create NAVI, supposedly, and are going with the older tech while MS gets the new tech, supposedly.
What reason does MS have to go after the crown? If your answer is that they need it to gain back market share, why can't sony's reason not be keep it or increase it further?
I'm still on denial stage and i'll keep hold of it till February.
And if it turns out to be true then i'll jump straight to acceptance.
PS5
36 CU / 2 GHZ
9.2TF
16 GB GDDR6 @ 530 GB/S
The rest of the info is just a map of how we got to it.
the difference is 1.8tf vs ~3tf. Assuming everything else equal I can absolutely see it.
Throw enough Alpha effects at the screen or increase your shadow resolution - watch that bandwith vanish.
Man,
I have no clue to believe. I choose to believe that both companies understand what they are doing.
Can you tell me which doc this is? I have only see 448... but I don't think I have found Oberon B0 docs yet.
User Banned (2 Days): Console wars
Microsoft probably thinks they need the power crown because of what happened this gen. I hope they realize the problem was that for years all Xbox fans had to look forwards to was cancelled games and Sea of Thieves.
Its a good attitude to have and honestly the correct one imo, There are no stupid people engineering consoles for these giant companies they know what they are doing and next gen all around is gonna be awesome i can feel it
I really think if they were using a RX 5700 it was just a place holder dev kit trying to get as close a possible to target specs. Klee source told him that both console are using better gpus than the one currently out. Which makes complete sense since the current gpu doesn't even support RT.
People thinking... series X will be hitting 12 navi teraflops are just dillusional...
its going to be two 9-10 tf machines battling it out... wich equals 12-13 navi tfs.
Phil never spoke of navi teraflops... he just said 2 times the performance of One X
its going to be two 9-10 tf machines battling it out... wich equals 12-13 navi tfs.
Phil never spoke of navi teraflops... he just said 2 times the performance of One X
They have 15 studios, most with multiple teams, along with a massive 1st party publishing division. Yes, they are all in on games.
look in the arden folder.
yes, the arden that is supposedly the anaconda chip. how did oberon get in there? how can the same amd engineer be working on two next gen consoles?
is arden oberon's final form?
Man, I bet you several Sony people are reading this thread laughing at all the Chicken Little going on in here. So have you all come to the conclusion that PS5 is 9.2TF? No take backs?
To me it just don't make sense when you can add few more CUs and lower the speed.
2Ghz is a insane speed .
By that logic, no amount of bandwidth is enough as you can always push past your allowed BW. ~50Gb/s per TF is "enough".
I know, but thats basically what Oberon is as far as we can tell by these test... an RDNA 1 Navi10 36CU clocked at 2 GHz with some improvements. Nothing about this is clear.
I mean this a speculation thread. People are allowed to change minds. New information is presented. Things change. This isn't some winner take all fight to death.
not flute, gonzalo. gonzalo had the firestrike score of 20000 up. whatever that means.
i went through a bunch of firestrike benchmarks and found many rtx 2080 firestrike scores under 17000 when paired up the cpu equivalent of the flute cpu. i even found 17000 scores for the rtx 2080 when paired with the ryzen 3700x.
dont put too much stock into those scores. my rtx 2080 has an overall score of 21.5k but ive seen some 5700xt scores over 30,000.
but yes, gonzalo was 1.8 ghz gpu, a 10% better gpu would probably push it closer to my 21.5k score. doesnt mean they are the same.
The spec of a machine is determined by the target price point and individual customisations. Everyone is working with the same suppliers & margins.
Your points regarding the GPU don't stand upto scrutiny. MS's next-gen titles have to match the capabilities of the 4TF Lockhart base model. In reality that means lower resolution and (for example) 'Standard' vs 'Ultra' settings.
You won't have a diminished experience as the CPU & SSD will be doing the heavy lifting. Improved AI, destruction, larger (more interactive) game world's. Developers won't be GPU limited. 3rd party epics like RDR2 & Cyberpunk where built with base XB1 & PS4 in mind. Machines with under 2TF. Mid-gen refreshes only gave us a resolution boost and didn't have exclusive titles.
Next-gen that baseline will be Lockhart at 4TF. Then we have PS5 at 9TF+ and XSX at 12TF. Given PS5 & XSX will both target 4K/60. And given developer's are going for a filmic look, therefore more open to resolution reconstruction techniques. Your guess is a good as mine as to what perceivable differences there will be in retail titles.
I'd suggest by the time devs start to really push these machines. We'll be into the mid-gen upgrade cycle. Believe me, however much these machines cost upfront. Everyone will be cashing in with mid-gen upgrades 3-4 years down the line.
Not having any technical knowledge at all, what are the odds of a dual gpu setup? As in, let's assume the Oberon specs are the PS5 GPU, tested all the way up to 2ghz.
The general consensus seems to be that this figure is very high and not realistic in a console (based on what we know at the moment).
So what does it look like if the clocks are much lower, but a dual gpu setup?
Are there ANY benefits at all in terms raw TFLOPS, heat, cost efficiency or ray tracing? Or is there no way Sony/MS move away from current APU approach?
The general consensus seems to be that this figure is very high and not realistic in a console (based on what we know at the moment).
So what does it look like if the clocks are much lower, but a dual gpu setup?
Are there ANY benefits at all in terms raw TFLOPS, heat, cost efficiency or ray tracing? Or is there no way Sony/MS move away from current APU approach?
Obviously. There's a difference between speculation and people steering conversation into one direction because they are so convinced that they know the answer when they don't know anything at all. Doesn't help when you have new posters coming in this thread all the time and people are chomping at the bit to tell them definitive info. It's muddling the conversation.
It's where my guess is at the moment yes, only because I feel it's grounded and I won't be disappointed if it turns out to be true.
Ok let me ask this since it was pointed out that Arden seems to have BC clocks for Xbox consoles. It was proposed that PS Now might not be utilizing PS5 hardware on Azure, what if Arden is a chip/whatever that's meant to be put in Azure and is meant to run Xbox and PlayStation games all on one package?
Possible?
//Arden_COM_BC1_228_0617/case_Noini.csv
testname,func pass,perf pass,Noini
cs_buffer_load_32bit_c2_cacheswizzle0_half_channel_arden,PASS,,0
cs_buffer_store_32bit_c2_cacheswizzle0_half_channel_arden,PASS,,0
cs_2buffer_op_32bit_c2_cacheswizzle0_half_channel_arden,PASS,,0
ray_tracing_tri,PASS,PASS,0
vrs_2x2_512_4x,PASS,PASS,0
hqmgg has said its Xbox stuff but he doesn't know much about it.:
#7,244
#7,248
AegonSnake by the way, he shows this many projects:
[introduce]
project = (Arden, Mero, mi100, Navi10, NAVI12, Navi12, Navi14, NAVI14, Navi21, Oberon, Renoir, Sparkman)
Edit 2: this "Arden folder" was within another folder that it looked like he was sending to his supervisor on May 6
Last edited:
It's okay, you are in the denial phase, don't worry, you'll get past that!
Which is why we shouldn't start making conclusions like some are doing in this thread.
For some reason when i quoted Burrman it messed up the ending quotes. So i wasn't saying that but Burmann was.
It's a navi rdna2 machine, as stated on MS's own website. Why would they say anything about GCN tflops? It's a navi rdna machine, any tflops listed are navi rdna2 tflops ffs.
I am with you, but I also like to dig and see at least what was being tested. This was 6-7 months ago. Could be different now.
Big Navi gpus might go to that length. We just have to wait for ces. I think big Navi can handle 2ghz plus we might see something bigger than 64cu count. 5700xt doesn't fit in the 12 tf calculation if it's rdna.
No need to deny anything since ill be getting both consoles at launch... Its just so that there is so little official information on what the specs are 100 percent.
And no one knows what realy is going on.
Yeah I think that this is the correct take. If it's higher, you're pleasantly surprised. If it is what it is, you're not disappointed. If it's lower, then somebody messed up.
You can believe in the 9-10 TF. We can believe in the 12 since we have info backing that. We'll see :)
yeah thats no problem... but doent go assuming anything sony since there is no information on that.
And all the news sites wich are quoting the insider information on the 9tf are damaging sony at this moment with no real cause to it.
ah ok. that makes sense.
arden was leaked to be navi 21 lite, 7nm+ and rdna 2.0 by a guy on twitter four days before the rogame leaks. sparkman is navi 12 lite and oberon is navi 10 lite according to that guy.
its not a conspiracy. ms simply paid more. a whole lot more to get the 7nm+ card.
- Status