-
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 |OT7| - nm
- Thread starter Mecha Meister
- Start date
- Status
So we are back at PS5 being more than 10TF, right? Yesterday we were in full panic mode at 8TF, so I'm expecting 13TF or higher by tomorrow.
We've literally seen the DS5 with the dev kit and there's no screen. Just a plain touchpad.
There's loads of reasons to doubt that rumour - but I think for legit rumours, whichever ones they are, any specs devs will have been quoted would have been based on what Sony or MS was telling them. And I doubt Sony or MS were going around quoting 'GCN flops' based on what was or wasn't around in the public domain re. RDNA. I mean for them it's always been Navi/RDNA, they've never not been dealing in that. This particular rumour is questionable for lots of other reasons though, IMO.
Once again, I am having trouble comprehending the quote of a quote chain. WHo is a dev and who is saying things about their relative power?
wait... Duke never said anything about a trusted source giving him the pastebin. Where's the post?
Pre-launch relative comparisons are by and large done on paper-specs alone, and that's never been subject to devkit availability. Though IME people don't 'leak' relative comparisons anyhow, that's usually done after the fact by whoever 'posts' the leaks (ie. not the source of information). I mean even this gen, it was DF (and others) analysis of complete dev-doc dumps that did all the comparing leading up to launch, not the devs ;)
And more tangential, but the real-world data isn't likely to be 'relevant' until several years in anyway - doing a recompile of your 10-15 years old codebase(which is how earliest stuff usually comes about) says more about your software stack than it does about new hardware.
I would be careful about comparing peak numbers for PC drives that are actually in the 100s of MB/s in more realistic workloads, to spec 'leaks' that haven't been given any context as of yet. While we don't know for sure, if these are numbers system guarantees developers (which would be a number much more relevant than peak bandwidth to know for a dev) the real-world speed comparisons could look very different.
And that's before factoring in the rumored hardware decompression which inflates effective read-speeds by 2-3x, something that just doesn't exist on PC consumer hardware at all (and you'd need a 3990x to brute force this effectively).
Reminder also that we're coming from drives rated for 20-40MB/s to the application, so it's a multiple orders of magnitude jump in any scenario.
Last edited:
thomasmahler and Matt are devs. I'm just saying target specs have been known by some devs for a while.
yea, if we are talking about comparing the specs with the assumptions we have, then only counting the likely cases, at worst the PS5 is 3.2Ghz CPU and 10.2TF GPU, both of those are withing margin of 10~15% difference.
10% slower CPU didnt do anything for the xbox one over the PS4.
15% slower GPU is comparable to 2160p to 2000p.
both are such small differences, that it really doesnt matter.
the only points to compare still would be the RAM and the SSD, Scarlett has 16GB, PS5's is unknown, but rumors point to either 16GB or 18GB, for scarlett the SSD bandwidth is mentioned as "gigabytes a second", for the PS5 we know its faster than consumer PCIE 3 SSDs, so at least 3.5GB/s.
I believe he was talking about the target specs.
Supposedly some devs reported than the target specs had been handled to them and there were rumours than the target specs for the Pa5 had a higher number of Tflops than Scarlett's.
Andrew Reiner mentioned he heard that same rumour from some developers and shared it in twitter just before the E3 show this year and there was a meltdown of angry fanboys in twitter that pushed him into not sharing any more details about.
Last edited:
One is an aim of intent from MS, so basically has nothing to do with what is actually true re performance.
Note tom warren also recently quoted Benji Sales saying PS5 is 13TF. Which is ~8% faster than 12TF, which is about the edge that most trusted sources state the PS5 has over Scarlett.
We learned that Scarlett is 12TF. Nothing more, and certainly nothing suggesting PS5 is suddenly 2TF the inferior of Scarlett when every credible source to this point suggests that the PS5 has a slight edge.
There are a lot of narratives being pushed in this thread that are nonsense and presented as being factual when they are anything but.
EDIT: Pastebin may or may not be true, but I would not put it above journalists and developers. Of note 10 vs 10.24 is not the edge we were told PS5 had which was closer to 10%. If it was a 2.4% edge I doubt they would be saying it was faster.
Last edited:
Tom Warren quoted Benji saying something around E3. Tom made no comment as to whether that was an accurate comment or not..
Oh yeah, look at me going super crazy! Like...what? You joked that he should be redeemed, I reminded you that he was a shit poster. In a fit of rage, apparently 😂
yep. Era's insiders won't tell us any more. Because they are Era readers, they don't want to get banned or mess the forum up.
Any more leaks, It will be someone from the outside.
People need to stop with the pastebin nonsense.
Also people talking about these specs like they are new there are not .
Because of Tom people only starting to thinking there possible when they were in front our face for months.
EDIT some people being pessimistic really limit what happen in this thread.
Also people talking about these specs like they are new there are not .
Because of Tom people only starting to thinking there possible when they were in front our face for months.
EDIT some people being pessimistic really limit what happen in this thread.
to provide context when people were mentioning 12-13TF?
It may imply it is true, but it doesn't mean it for certain. And as this is a speculation thread, and we have had more concrete 'confirmation' from Tom on Anaconda, I just think it'd be worth holding off on considering it a confirmation until it comes from him directly
MS is reacting realtime on changes within Sony or is aware what is happening at Sony. I would have expected Sony to have the upperhand considering the big succes with PS4.
- Old PS5 > 8TF
- Old X > 10TF
New PS5> 10.7+ TF
New X > 12+ TF
Cant deny MS is making sure there is a gap no matter how small.
- Old PS5 > 8TF
- Old X > 10TF
New PS5> 10.7+ TF
New X > 12+ TF
Cant deny MS is making sure there is a gap no matter how small.
It woudln't even remotely work like that. Both companies would be iterating on targets as the technology matured. And in the final stages it shouldn't be a surprise to see MS eg tweaking clock speeds to get more power - they did the same on XB1.
Not saying it is true. I just find it interesting that people jump on the leak that scarlett is 12TF as meaning PS5 must be less. Even though we have literally nothing outside of one pastebin to suggest so, and all other rumours suggest that PS5 has an edge.
Like I said it's interesting the narratives people like to push.
We learned that Scarlett is 12T but nothing new regarding the relative power of Scarlett and PS5.
Regarding what matt said recently (last 3 posts in next gen thread) I would not say it corroborates that ps5 is 15% more powerful, matt just gives the figures, does not say which platform those figures favour.
What is funny is we know that Scarlett is 12T for months but hardly any one took that old leak seriously .
Tom just made some people open there eyes that some of that is possible.
12TF puts Scarlett at 56CU @ 1675Mhz
If Sony really do have dual sided cooling compared to Scarlett's one sided cooling (based on the Sony patent) eking out another 125Mhz wouldn't be that hard. Especially if Scarlett is clocking its Zen 2 @ 3.5Ghz compared to 3.2Ghz on the Sony APU.
The fact we know that MS are going for 12TF with a 3.5Ghz Zen 2 shows that a 12.9TF GPU with a 3.2Ghz CPU is very possible.
Unless the Windowscentral leak is not true that is.
If it is true then a 12.9TF PS5 is very, very possible. And actually unifies the Windowscentral leak with the rumours regarding the power differential between the PS5 and Scarlett.
My money is on Scarlett having 3.5Ghz Zen 2 and 56CUs @ 1675Mhz with PS5 having 3.2Ghz Zen 2 and 56CUs @1800Mhz. Sony have 10% better GPU performance, MS have 10% better CPU performance.
Last edited:
IMO team "realistic" (aka debbie downers) have held this thread back since the beginning.
Everything that has turned out to be true had discussion shut down because it was "impossible".
Edit I misread a post of matts so here is something devs may be using a lot next gen
Last edited:
Colbert held it back with his predictions, experiments etc. He has even an avatar bet with someone :) I dont wont to call him debbie downer (seems harsh) but he is a party pooper. :)
Maybe I read that wrong but I'm pretty sure he stressed this wasn't a hint at actual gaps but was just an example of what might be less substantial (to him) vs more substantial (20+%). I don't think we should start saying 'he said the difference is 15%' if that's the case.
He didn't say the difference is 15% so the rest of your post is conjecture based off misquoting someone.
This thread really moved. Will be interesting to see where the consoles are at when the final specs are revealed.
#teamrealistic may be right though but Phil and cerny have just lost the plot and are unhinged on there quest for MOAH POWAAAA..
AT 12 TF these beasts are going to be drawing like 250watts!
AT 12 TF these beasts are going to be drawing like 250watts!
Tom isn't supporting this rumor as fact. He has stated that. He was just pointing to why he made that tweet 6 months ago.
he knows the actual numbers but is holding off.
I'm pegging between 9.5 - 10.5 on PS5. I really think they went with the 40 CU Oberon leak and have been adjusting the clocks quite a bit to hit cooling.
40 CU makes PS5 back compat available entirely with disabling CU and changing clocks which lines up perfectly with the Sony patent on it.
This also can put them in the $399 camp. Considering how well that worked for them last time I can imagine they really want to hit it again.
Lockhart rebirth was likely driven by this knowledge
He said in this thread today that he never said the 13 TF was true or not. Was just passing along what the rumor was.
- Status