Support NeoGAF

[CuriousToe]

Some people like repeatedly reading the old tech spec PDFs of chip parts online and bringing it up at every opportunity. Others like reading Game of Thrones. They're both good but one's definitely more fictional than the other

Fixed.

 

[Self High Five]

Still talking about github.

 

[thicc_girls_are_teh_best]

Fixed.

December 2019 and January 2020 must be ancient eh?

I was only replying to the stuff Zero707 linked because I found it interesting. Whether you like it or not, it's still highly relevant to the discussion, and that doesn't look to be changing for a very long while.

So if we're gonna try to figure things out it's better to include it than exclude it, that's just going from one extreme to the other.

 

[Sosokrates]

Out of all those "leaks" most likely only about 10% of them are accurate.

Even before this gen there was "verified insiders" who turned out to be completely wrong.

 

[01011001]

everyone on here who dismisses github is literally in absolute denial mode.
you can think it might be only part of the picture, but it is accurate data.

also going with less CUs and super high clocks can actually be beneficial in some areas. meaning even tho the PS5 would have less TFLOPS it could still have advantages over the SX because of the high clock rate

 

[Gamernyc78]

everyone on here who dismisses github is literally in absolute denial mode.
you can think it might be only part of the picture, but it is accurate data.

also going with less CUs and super high clocks can actually be beneficial in some areas. meaning even tho the PS5 would have less TFLOPS it could still have advantages over the SX because of the high clock rate

 

[Gavin Stevens]

It’s not that GitHub is wrong, far from it. It’s that github is used in this case as a repository for test data, and without actual knowledge of what the product is, that’s worthless. See they could actually hve an entry labelled “PlayStation”, but that wouldn’t mean it’s the spec they will use, essentially because every company on that scale always has new things in R&D, even stuff that will never see the light of day, just to see how things work in a set environment.

Github is fine. People saying it’s not fine based solely on MS owning it are the ones that are crackers. Can you imagine the lawsuits they would be open to?

 

[FERN]

It’s not that GitHub is wrong, far from it. It’s that github is used in this case as a repository for test data, and without actual knowledge of what the product is, that’s worthless. See they could actually hve an entry labelled “PlayStation”, but that wouldn’t mean it’s the spec they will use, essentially because every company on that scale always has new things in R&D, even stuff that will never see the light of day, just to see how things work in a set environment.

Github is fine. People saying it’s not fine based solely on MS owning it are the ones that are crackers. Can you imagine the lawsuits they would be open to?

What would your guess be for what it is? BC compatibility testing for PS4?

 

[01011001]

only because of you I hope github shows final specs... literally just so you choke on your childish responses.
I don't think they are, but I hope, just because of you and your peers I would fucking fall of my chair laughing... again, low chances but I hold on to that small possibility just for you

 

[WhispersintheWind]

This is total bollocks. There is no way Sony would use a 512-bit GDDR6 memory bus. As far as I am aware, the memory traces need to be very carefully managed with GD6, which would be appallingly difficult with a 512-bit bus. Not to mention the fact that GD6 memory controllers are fucking massive, and having 8 64-bit controllers would take up a stupid amount of die space. Given all this, they would quite literally be better off using HBM2 and using the die space on a larger GPU. Absolutely crazy.

512 Bit Memory Bus wouldn’t be enough to keep the GPU fed. So wouldn’t be no point of having 15TF.

 

[Gavin Stevens]

What would your guess be for what it is? BC compatibility testing for PS4?

Portable? A new phone? A watch? Maybe ear rings that when you shake them, they make a screen fold out and play games?

Like I say, without context, it could be anything. It’s good to make assumptions based on known candidates, but never forget that just because you have a 4tf bread toaster in your hand, it likely had about 50 prototypes at some stage all of various designs and features and combinations. You never just make one and done systems, and you always do R&D, even long after the system has finished and been released.

 

[Hatsuma]

join us brother

Idfk why,but this post has me in tears. That pic with the caption. Must be the gummy...

 

[Disco_]

Wtf dude your past 2 pages are non stop posts about github.are we back to 3 months ago again ? God hope not.

It's always funny when a new account shows up and immediately posts with authority and knowledge. No warming up to the place, not taking it easy. Just picking right up where their last account left off...

 

[pawel86ck]

everyone on here who dismisses github is literally in absolute denial mode.
you can think it might be only part of the picture, but it is accurate data.

also going with less CUs and super high clocks can actually be beneficial in some areas. meaning even tho the PS5 would have less TFLOPS it could still have advantages over the SX because of the high clock rate

There was a time when I was considering github leak (and some people here even attacked me for that), but given what we already know about RDNA2 and new cooling patent 9.2TF looks more like a joke to me and no sane persone should still consider it. PS5 is either close to 12TF XSX, or even faster like Osiris Black suggest.

 

[devilNprada]

Github is fine. People saying it’s not fine based solely on MS owning it are the ones that are crackers.

I am not familiar with the circumstances... Again I ask... Did the same person who was an intern test both the MS chip and the Sony chip?
If that's the case; doesn't anyone else think that odd?

 

[thicc_girls_are_teh_best]

only because of you I hope github shows final specs... literally just so you choke on your childish responses.
I don't think they are, but I hope, just because of you and your peers I would fucking fall of my chair laughing... again, low chances but I hold on to that small possibility just for you

Yeah, the dude you quoted is a complete clown, it would be worth it just for their sole disappointment if nothing else.

But that said, I don't actually "want" that to happen. I don't really "want" anything in terms of these system specs honestly, they'll turn to be what they are and that's that. I can see Gavin Stevens 's point and he's more or less right, but we can extend that to the "leaks" and most of the rumors as well. Given the vagueness of majority of rumors, it's safe to say most of the insiders don't know what the PlayStation 5 as a whole product actually is, either, but I wouldn't say the more realistic rumors are useless. They're just missing context.

All the same I do just have to acknowledge that the information pertaining to Oberon and such has been very consistent, there's a code to deciphering it and a timeline for it as well. And while they haven't provided full context, they've been gradually providing a good deal of it. Why would anyone want to throw that out of the window? And again like Gavin Stevens was saying, people trying to dismiss it because MS has a stake in Github is ridiculous, because of the lawsuits.

I don't see why MS would be spreading FUD about a potential (and very key/crucial) ally to their cloud services when they know they'll need as much of that goodwill with Sony (and Nintendo) as possible to combat the likes of Google, Amazon etc. in that area. It would be completely counter-productive not to mention unprofessional.

 

[FERN]

There was a time when I was considering github leak (and some people here even attacked me for that), but given what we already know about RDNA2 and new cooling patent 9.2TF looks more like a joke to me and no sane persone should still consider it. PS5 is either close to 12TF XSX, or even faster like Osiris Black suggest.

What new cooling patent?

 

[Gavin Stevens]

I am not familiar with the circumstances... Again I ask... Did the same person who was an intern test both the MS chip and the Sony chip?
If that's the case; doesn't anyone else think that odd?

Chip testing can “sometimes” be outsourced, and if you’re dealing with multiple variants of a single product that has multiple instances with varying elements and conditions, you will outsource those to different companies to test performance.

But honestly, who knows. It’s like if I was to ask you what my wife is doing now, could you tell me? No, because I don’t even know. All I know is she’s been in the bath an hour and for all I know she’s succumbed to corona and drifted off peacefully... which would mean I get the whole bed to myself tonight. So, swings and round abouts...

 

[Zero707]

It's always funny when a new account shows up and immediately posts with authority and knowledge. No warming up to the place, not taking it easy. Just picking right up where their last account left off...

new account...........

 

[01011001]

There was a time when I was considering github leak (and some people here even attacked me for that), but given what we already know about RDNA2 and new cooling patent 9.2TF looks more like a joke to me and no sane persone should still consider it. PS5 is either close to 12TF XSX, or even faster like Osiris Black suggest.

we only know Ariel was RDNA1 that is not confirmed for Oberon, meaning RDNA2 is not really a factor of authenticity here, also that cooling patent suggests high clock speeds... also very much in line with the 2GHz on Github, So I really don't see how this is of any real concern in this case.
as long as we don't have new details that aren't second hand knowledge it's hard to tell what has actually changed since the latest found tests, or how much of the chip was used.
insider infos range wildly up and down, so where should we look to and what should we ignore?

github isa point of information that we at least know is accurate, we just don't know how much it represents final hardware and what what's the full context of the tests. that's where the speculation should revolve around.
we have this data, what should we make of it? this is the real question here

 

[thicc_girls_are_teh_best]

512 Bit Memory Bus wouldn’t be enough to keep the GPU fed. So wouldn’t be no point of having 15TF.

It's speculated to be a 256-bit memory bus right? Because in that case it definitely doesn't add up as a rumor. Then again some of these latest Reddit/4Chan/Pastebin rumors seem to come from people who don't understand how this tech works

 

[01011001]

Yeah, the dude you quoted is a complete clown, it would be worth it just for their sole disappointment if nothing else.

But that said, I don't actually "want" that to happen. I don't really "want" anything in terms of these system specs honestly, they'll turn to be what they are and that's that.

same, all I always said here is that, we got this data, don't ignore it, try to make sense of it, don't jump on any freaking leak just because you like it and don't expect too much.
somehow this is controversial... because if you say something along these lines you're an Xbox fanboy... even tho lately I often have to recommend to not play certain games on Xbox but to get them on PC instead because the Xbox version sucks (most recently Ori 2) and I dod that on this very forum... what a very Xbox fanboy thing to do isn't it? lol

 

[Zero707]

if I was to ask you what my wife is doing now, could you tell me? No, because I don’t even know. All I know is she’s been in the bath an hour and for all I know she’s succumbed to corona and drifted off peacefully... which would mean I get the whole bed to myself tonight. So, swings and round abouts...

 

[Mod of War]

Everybody calm the hell down and quit troll baiting each other like little kids. You know who you are.

 

[pawel86ck]

What new cooling patent?

I havent seen anything like that before in any console, and I bet cooling from two sides will provide supperior results (and very high GPU clock) even compared to XSX.

 

[Zero707]

Everybody calm the hell down and quit troll baiting each other like little kids. You know who you are.

 

[thicc_girls_are_teh_best]

same, all I always said here is that, we got this data, don't ignore it, try to make sense of it, don't jump on any freaking leak just because you like it and don't expect too much.
somehow this is controversial... because if you say something along these lines you're an Xbox fanboy... even tho lately I often have to recommend to not play certain games on Xbox but to get them on PC instead because the Xbox version sucks (most recently Ori 2) and I dod that on this very forum... what a very Xbox fanboy thing to do isn't it? lol

Yeah I heard a couple people mention Ori 2 has a few framerate issues on Xbox, surprising if true. Maybe they can fix it in a patch. Seems to be very liked tho; it's nice seeing creative games like Ori 2 and Dreams being critically praised even if they aren't gangbusters in terms of sales.

But to the rest of the point tho, yeah unfortunately that seems to be very popular in general these days. I blame the hyper-partisan political/cultural effect. Everything either has to be completely to one's agreement or else it's the enemy and must be destroyed. Complete binary, us vs. them bullshit. Critical nuance and weighing of options even if they mean compromising on a few of your ideas on the subject to get a probable outcome/solution just seems to be a dying art.

 

[Batiman]

Good news for PS only guys. 2 of the best games this gen. Loving the hell out of wisps.

 

[Sosokrates]

If sony go only powerful and price $499 , Lockhart will take some market share from them.

 

[01011001]

Yeah I heard a couple people mention Ori 2 has a few framerate issues on Xbox, surprising if true. Maybe they can fix it in a patch. Seems to be very liked tho; it's nice seeing creative games like Ori 2 and Dreams being critically praised even if they aren't gangbusters in terms of sales.

But to the rest of the point tho, yeah unfortunately that seems to be very popular in general these days. I blame the hyper-partisan political/cultural effect. Everything either has to be completely to one's agreement or else it's the enemy and must be destroyed. Complete binary, us vs. them bullshit. Critical nuance and weighing of options even if they mean compromising on a few of your ideas on the subject to get a probable outcome/solution just seems to be a dying art.

well on One X they could iron out most of the issues I guess, but looking at DF's video, that base Xbox One version REALLLLY looks rough... like almost unplayable levels of rough. it goes down into the 20s! that's fucking crazy.
I started playing it on One X and very early was like "nope... OFF TO THE PC TO DOWNLOAD IT AGAIN!"
and since I now have the PC connected to the same TV anyways (10m HDMi cable + USB extender cable with plugged in bluetooth dongle lol)... I mean... doesn't really have much of a negative impact (it is a bit of a hassle to launch it with a multi screen setup, but a few seconds of hassle for a much smoother game is worth it IMO)

 

[thicc_girls_are_teh_best]

well on One X they could iron out most of the issues I guess, but looking at DF's video, that base Xbox One version REALLLLY looks rough... like almost unplayable levels of rough. it goes down into the 20s! that's fucking crazy.
I started playing it on One X and very early was like "nope... OFF TO THE PC TO DOWNLOAD IT AGAIN!"
and since I now have the PC connected to the same TV anyways... I mean... doesn't really have much of a negative impact (it is a bit of a hassle to launch it with a multi screen setup, but a few seconds of hassle for a much smoother game is worth it IMO)

Well well, almost sounds the didn't let the base XBO hold back the game on the X. It's almost as if this could serve as a parallel example for XBO and another system (maybe) launching this year xD...

Spoiler

I kid; people are gonna find a way to argue XSX being held back by XBO anyway. The power of delusion demands it

That said, ouch to anyone playing it on XBO and they shouldn't be having those issues on the X. So it sounds like at the very least they need to optimize the X version and release the fix as a free patch.

 

[CyberPanda]

 

[3liteDragon]

Quick question, what are the chances of the PS5 having a dual-GPU?

 

[pawel86ck]

we only know Ariel was RDNA1 that is not confirmed for Oberon, meaning RDNA2 is not really a factor of authenticity here, also that cooling patent suggests high clock speeds... also very much in line with the 2GHz on Github, So I really don't see how this is of any real concern in this case.
as long as we don't have new details that aren't second hand knowledge it's hard to tell what has actually changed since the latest found tests, or how much of the chip was used.
insider infos range wildly up and down, so where should we look to and what should we ignore?

github isa point of information that we at least know is accurate, we just don't know how much it represents final hardware and what what's the full context of the tests. that's where the speculation should revolve around.
we have this data, what should we make of it? this is the real question here

You still have dobuts PS5 will use RDNA2 after AMD presentation? RDNA2 is a given and if github 9.2TF suggest RDNA1 it only proves it was very old SoC (we know Sony wanted to launch 8-9TF PS5 in 2019).

 

[Panajev2001a]

More GitHub leak based rumours and speculation ?

 

[CyberPanda]

Quick question, what are the chances of the PS5 having a dual-GPU?

Not happening. Dual GPU would cause a headache for developers.

 

[ANIMAL1975]

we only know Ariel was RDNA1 that is not confirmed for Oberon, meaning RDNA2 is not really a factor of authenticity here, also that cooling patent suggests high clock speeds... also very much in line with the 2GHz on Github, So I really don't see how this is of any real concern in this case.
as long as we don't have new details that aren't second hand knowledge it's hard to tell what has actually changed since the latest found tests, or how much of the chip was used.
insider infos range wildly up and down, so where should we look to and what should we ignore?

github isa point of information that we at least know is accurate, we just don't know how much it represents final hardware and what what's the full context of the tests. that's where the speculation should revolve around.
we have this data, what should we make of it? this is the real question here

we only know Ariel was RDNA1 that is not confirmed for Oberon,

as long as we don't have new details that aren't second hand knowledge

First hand knowledge, you welcome:

True leaks are always true at a certain time and circumstances.

Today, 9.2TF, is the only thing that PS5 does not have.

Now is at puberty.

 

[Agnostic2020]

Not happening. Dual GPU would cause a headache for developers.

Not only that but its expensive. Why not add 20 more cu for 35$ instead of doubling the cost ?

 

[FranXico]

The main issue with dual GPUs is not hindrances to developers. The SDK could provide an abstraction to hide the split from them.
The biggest problem with a dual GPU is that it is expensive as hell and complicates the chip design for no good reason.

 

[THE:MILKMAN]

Interesting he goes with the one spec (36-40CUs/9.2TF) all insiders seem to agree it isn't.

Lets see how this plays out.

 

[3liteDragon]

I havent seen anything like that before in any console, and I bet cooling from two sides will provide supperior results (and very high GPU clock) even compared to XSX.

With this patent (if it ends up happening) and RDNA 2's efficiency combined, a 2 GHz clock makes even more sense now than it did back then when the GitHub leak was first discovered. And on top of that, we don't even know what kind of "advanced" cooling solution Sony's even using for the PS5, excluding the double heatsink patent.

 

[it_wasn't_me]

More GitHub leak based rumours and speculation ?

Oh snap, software raytracing. LOL

 

[thicc_girls_are_teh_best]

First hand knowledge, you welcome:

Don't know if it's been clarified if that info is reflective of Ariel iGPU profile or not. Anyhow it is a very early revision of Oberon, and later revisions have already switched to higher-clocked chips at higher memory bandwidth for the bus.

The Navi 10 listing there could be pointing to a number of things and in light of more recent data its context is muddled. But we should probably be working on how to get that data to agree with a forming POV instead of dismissing it; insider rumors lack a lot of context themselves (to just one example; almost all of them only provide very basic specs like TF count, when they should have access to much more info on the systems than that if they have close connection to particular sources, IMHO), but we don't use one or two oddities to throw the rumors completely out do we?

 

[Moridianae]

I'm still in that 11-12TF range. Perception is reality. Doesn't matter if we know other components even the playing field.

If the layman or general gamer perceives the literal face value difference of 9.2 as less than 12, they see it as weaker. That is the reality of it.

That is why I see the 11.6 being plausible..they just round it up to 12 and call it a day for good PR, because it is essentially "even" with the XSX at face value.

They go 13TF and the same situation arises with good PR, because the literal face is it's "better", even though we guru's know that's not the case.

The *only* way I see them going 9.2TF is a $399 price tag. That being said, $399 does NOT mean it HAS to be 9.2TF. No one knows what their true BOM is, what their cushions are, what loss they are prepared to take, nothing.

 

[FranXico]

It’s not that GitHub is wrong, far from it. It’s that github is used in this case as a repository for test data, and without actual knowledge of what the product is, that’s worthless. See they could actually hve an entry labelled “PlayStation”, but that wouldn’t mean it’s the spec they will use, essentially because every company on that scale always has new things in R&D, even stuff that will never see the light of day, just to see how things work in a set environment.

Github is fine. People saying it’s not fine based solely on MS owning it are the ones that are crackers. Can you imagine the lawsuits they would be open to?

I think I can speak for most people in here when I say that when they write "the GitHub data is wrong", they actually mean "that interpretation of the GitHub data is wrong", but are keeping it short.

 

[alex_m]

There is 0% chance that the final retail version of the PS5 is 11.6tf if the XSX really is a 12tf machine. Make no sense from a marketing point of view.

But you know, some people need to be believe that, for once, MS is going to release the most powerfull console in the market at launch. They need that pyrrhic victory

As I understand the leak, RT module will take care of ray tracing, while those CU's do the rendering of the scene. In AMD marketing terms it would be like 15+ SecretSauce TF.

 

[Bo_Hazem]

This guy is already struggling while he speaks, he should be careful not to get coronavirus as he would suffocate. Probably he's struggling that way because of all the BS he has to throw through his mouth.

 

[Gudji]

There is 0% chance that the final retail version of the PS5 is 11.6tf if the XSX really is a 12tf machine. Make no sense from a marketing point of view.

But you know, some people need to be believe that, for once, MS is going to release the most powerfull console in the market at launch. They need that pyrrhic victory

EDIT: nevermind I'm almost sleeping. Yeah you're right, if its 11.6 they're going to say it's 12 (at least on stage).

 

[Neofire]

I seen redgamingtech's video a few minutes ago and I'm perplexed at how Sony can squeeze more performance out of a 36-40 CU GPU with "magic" customizations then AMD can out of the the same cards with those exact amounts of CUs at around 8-9 TFs of raw power but will somehow will still compete with Microsoft's 12TF console?

I don't get it

 

[thicc_girls_are_teh_best]

This guy is already struggling while he speaks, he should be careful not to get coronavirus as he would suffocate. Probably he's struggling that way because of all the BS he has to throw through his mouth.

C'mon he's good people and they make some fun content. I don't think he deserves that level of dismissal (or a coy coronavirus jab either. Probably a little too fresh to do that given how easy it seems to catch and the deaths that've happened so far).

I seen redgamingtech's video a few minutes ago and I'm perplexed at how Sony can squeeze more performance out of a 36-40 CU GPU with "magic" customizations then AMD can out of the the same cards with those exact amounts of CUs at around 8-9 TFs of raw power but will somehow will still compete with Microsoft's 12TF console?

I don't get it

Because it's about more than just the narrow-minded obsession with teraflops. CPU, cache levels/speeds, the SSD , possible persistent memory, GB per TF ratio, OS optimization/utilities, OS resource footprint, dev team talent/size/time/budget resources, etc.

It all adds up, there's so much more to consoles than raw TF count.