Support NeoGAF

[Mass Shift]

If Sony or Microsoft skips this E3 then that will be a big mistake. Could matter in fact end up like SEGA

You could still have your press conference and skip E3 altogether.

 

[Mass Shift]

Probably... the "Oberon" only shows BC mode tests... I did not see Arden or Sparkman are these BC tests or full tests?

Honestly no way to know at this point. That's why I was asking.

 

[The Mad Draklor]

My question is, if Series X does have more TFs, why do you think some insiders are saying PS5 has the power edge?

It could be that they're equally powerful, but Microsoft has decided to upclock the GPU by 100mHz. Either way, I don't expect any discernible difference.

 

[ethomaz]

There is no way to make predictions without thinking the inside info if you have them lol

 

[Hawking Radiation]

I am afraid its going to be loud.

That goes against what Sony has been saying about PS5 going to be a 'premium' console ALSO,

Main complaints from a whole heap of PS4 owners is noise. Then you also have the know issue of shockingly low or poor dual-shock 4 battery life.

It would be incredibly daft for Sony to ignore the above complaints from current PS4 owners and push ahead with a new console with the same issues.

Sorry, i don't buy it. The PS5 will absolutely not be a loud console.

 

[David Ricardo]

Please don't start this topic again.

It wasn't my intention to start that topic again, cause I think it's now clear that all insiders and Phil retweeting the DF video that points to 12 TF RDNA, make obvious that Series X is at least 12 TF RDNA.

My point was that, to the best of my knowledge, kleegamefan was the first insider to clarify it was 12 TF RDNA.

 

[ethomaz]

Honestly no way to know at this point. That's why I was asking.

I'm trying to find the "leaks" again but I can't.
If you have the link of these github data to share it will be cool to have a look again.

PS. I can still find the Komachi modes where the clock match PS4 and PS4 Pro but nothing else.

 

[Dolodolo]

There is no way to make predictions without thinking the inside info if you have them lol

Our Video Game Predictions For 2019

What’s coming for video games in 2019? What sort of wild predictions do we have for the year? Today on Kotaku Splitscreen, we discuss.

kotaku.com

 

[ethomaz]

Our Video Game Predictions For 2019

What’s coming for video games in 2019? What sort of wild predictions do we have for the year? Today on Kotaku Splitscreen, we discuss.

kotaku.com

How do you know he not take inside info to make the predictions? He knows the inside info so he trend to make predictions that doesn't disagree with it.

Let's give an example:

- You have inside info A will launch in November 2020
- You make a prediction saying A will launch in Q1 2021? No, he will predict say A will launch in Q4 2020.

 

[johnjohn]

How do you know he not take inside info to make the predictions? He knows the inside info so he trend to make predictions that doesn't disagree with it.

He didn't get one 2019 prediction right.. They're not allowed to use inside info, so I'd say it puts Jason at a pretty big disadvantage when making these predictions lol.

 

[Dolodolo]

How do you know he not take inside info to make the predictions? He knows the inside info so he trend to make predictions that doesn't disagree with it.

Think what you want
none of his forecasts for 2019 came true
They have conditions. Without internal information
I don't think Jason Schreier is lying to his colleagues

 

[ethomaz]

Think what you want
none of his forecasts for 2019 came true
They have conditions. Without internal information
I don't think Jason Schreyer is lying to his colleagues

So he makes wrong predictions on purpose even knowing it is wrong?
What the point?

Or

He only make predictions about thinks he doesn't have inside info?
Boring predictions but yeap he did not use inside info.

 

[johnjohn]

So he makes wrong predictions on purpose even knowing it is wrong?
What the point?

Or he makes predictions on things he as no inside info for. I don't think Jason is omniscient.

 

[DJ12]

My question is, if Series X does have more TFs, why do you think some insiders are saying PS5 has the power edge?

Thats an easy one.

When MS go RTX on the performance takes a hit, the better/discreet solution on PS5 performs better, all '9' of it's measly teraflops just for geometry, texturing and shaders all RT handled by the RT integrated chip. That '10%' difference soon disappears in this scenario.

That would be my guess anyway.

(play on word here, I have no insider info unlike Walt)

 

[johnjohn]

Thats an easy one.

When MS go RTX on the performance takes a hit, the better/discreet solution on PS5 performs better, all '9' of it's measly teraflops just for geometry, texturing and shaders all RT handled by the RT integrated chip. That '10%' difference soon disappears in this scenario.

That would be my guess anyway.

(play on word here, I have no insider info unlike Walt)

That could explain a performance edge, but Insiders like Klee have been clear that PS5 has more TFs.

 

[PocoJoe]

It could be that they're equally powerful, but Microsoft has decided to upclock the GPU by 100Hz. Either way, I don't expect any discernible difference.

0.0001MHz would not make a difference.

 

[saintjules]

 

[David Ricardo]

The "5" looks weird. Not digging it.

 

[The Mad Draklor]

0.0001MHz would not make a difference.

Sorry, 100 Mhz.

 

[sonomamashine]

Thats an easy one.

When MS go RTX on the performance takes a hit, the better/discreet solution on PS5 performs better, all '9' of it's measly teraflops just for geometry, texturing and shaders all RT handled by the RT integrated chip. That '10%' difference soon disappears in this scenario.

That would be my guess anyway.

(play on word here, I have no insider info unlike Walt)

I think that would draw more power than XSX big APU.

 

[Jtibh]

Wow this thread got sprinkled with some heavy feminist ingredient.
HATRED.

I am almost wishing next gen would start 2021 .
Whats happening here is too much for my little hart to handle.
Fake leaks , speculations, lies, fights.......

 

[DJ12]

That could explain a performance edge, but Insiders like Klee have been clear that PS5 has more TFs.

Maybe it will have more tflops, this is just based on the info now as that is what the comments are based on, current or fairly recent devkits.

I don't think Sony devkits are up to spec yet MS are probably sending the out with retail APUs at this point. (I don't know would be cool if someone could let us know the state of the devkits - Sony usually only get the powerful last revision 'retail' devkits out after the consoles have been released historically)

 

[Handy Fake]

Wow this thread got sprinkled with some heavy feminist ingredient.
HATRED.

I am almost wishing next gen would start 2021 .
Whats happening here is too much for my little hart to handle.
Fake leaks , speculations, lies, fights.......

... Pardon?

 

[DaGwaphics]

Wow this thread got sprinkled with some heavy feminist ingredient.
HATRED.

I am almost wishing next gen would start 2021 .
Whats happening here is too much for my little hart to handle.
Fake leaks , speculations, lies, fights.......

You'd better stay off psn or live. Yikes!

 

[Rossco EZ]

that looks shiteeeee

 

[Reindeer]

Wow this thread got sprinkled with some heavy feminist ingredient.
HATRED.

I am almost wishing next gen would start 2021 .
Whats happening here is too much for my little hart to handle.
Fake leaks , speculations, lies, fights.......

Welcome to the wonderful world of console gaming .

 

[Bladed Thesis]

He didn't get one 2019 prediction right.. They're not allowed to use inside info, so I'd say it puts Jason at a pretty big disadvantage when making these predictions lol.

Because he's a fucking moron.

 

[ManaByte]

Because he's a fucking moron.

When you decide to die on the hill being a PR mouthpiece for a video game forum, that's a given.

 

[Bladed Thesis]

When you decide to die on the hill being a PR mouthpiece for a video game forum, that's a given.

It's not even that. He think so highly of himself and completely unable to take criticism or comments on his writing. He acts like his book is god's gift to gamers. He acts like his expose pieces are Watergate 2.0. The guy is sniffing his own farts to a massive degree.

 

[DeepEnigma]

When you decide to die on the hill being a PR mouthpiece for a "video game" forum, that's a given.

That's better.

 

[DJ12]

I think that would draw more power than XSX big APU.

Don't know how you could make that assumption baring in mind we know absolutely nothing about Sony's raytracing tech other than in December 2018 they stated they've been working on it for a while and were still working on it.

Anyway, that's the concern of my electric company and the bank manager.

For those that haven't really seen the tech in action (I hadn't to be fair. I just turn RTX on on my PC)



This is fully pathtraced (AKA ray traced) and while the frame rate noticeably takes a hit at times, it's pretty impressive.

This is not what AMDs solution offers although the results will probably look comparable.

 

[Rightisright]

He said twice the performance. That would be less than 12 tf RDNA. Kleegamefan made his RDNA clarification before the first DF video, which guessed Phil could be talking about 9-10 RDNA teraflops which are about twice the performance of One X.
Obviously, in specs given to devs, they would only use actual teraflop numbers.


In a world where everybody says the truth, you wouldn't need to double check anything. Regrettably, we are not living in that kind of world. Also, if several insiders have the same specs, it means that they can't be tracked by Microsoft just by looking at the leaked numbers.

He didn't say twice the performance, he said twice the power. Two different things.

“So when we do the math, we’re over eight times the GPU power of the Xbox One, and two times what an Xbox One X is.”

 

[thicc_girls_are_teh_best]

And let's be honest crowing about power when we know Sony has at least two things (much faster ssd and better ray tracing) that could make up the deficit is probably not the smartest idea at this time.

These aren't absolutes; just speculation based on certain insiders. That said, if they are true, they will certainly give PS5 an advantage in those areas, but they are not exactly equivalent to teraflop performance or CPU performance; none of them can be directly compared to one another, because they have their different use-cases and areas where they excel.

Interesting that Jason Shreier predicts Sony will skip E3 again this year.

https://kotaku.com/our-video-game-predictions-for-2020-1840903755

I want whatever ganja Schreier is smoking because that would be absolutely crazy to skip E3 twice in a row. Hope that wouldn't be a sign of anything like a delay :S

Maybe it will have more tflops, this is just based on the info now as that is what the comments are based on, current or fairly recent devkits.

I don't think Sony devkits are up to spec yet MS are probably sending the out with retail APUs at this point. (I don't know would be cool if someone could let us know the state of the devkits - Sony usually only get the powerful last revision 'retail' devkits out after the consoles have been released historically)

I thought Sony was supposed to be ahead in devkit progression according to a lot of the insiders, this would contradict that point.

 

[Disco_]

The 1060 had 980 like performance, yet it retailed even at 230$. It is not unheard for mid range next gen cards to match high end last gen cards.

Until the 2000 series cards, that was the norm for NVIDIA. With no competition from AMD, they hiked margins, reduced performance gains for lesser tiers.

So when Arden and Sparkman were stress tested, full CU and RT were enabled?

In the github docs? Correct me if I'm wrong but neither sparkman nor arden had any measured results in any of the github docs.

 

[Disco_]

I want whatever ganja Schreier is smoking because that would be absolutely crazy to skip E3 twice in a row. Hope that wouldn't be a sign of anything like a delay :S

You should consider the context of his predictions. He is NOT ALLOWED to use any of his insider knowledge when making these predictions. Hence they're all mostly based around things he has no info about. None of this 2019 predictions turned out to be true. That should tell you how accurate his 2020 predictions will be.

 

[ethomaz]

In the github docs? Correct me if I'm wrong but neither sparkman nor arden had any measured results in any of the github docs.

Do you have a link with that doc?

 

[it_wasn't_me]

We can already throw out the idea that both sides are not going for smart designs and one is simply going for brute power. The way it sounds like Microsoft has implemented their SSD to assist their GPU, using it like a sort of virtual ram that may be connected pretty close to the GPU itself, is the very essence of smart design and not simply just relying on brute power. And rumors suggest it is Sony that's pushing their GPU to 2GHZ+ (a speed we know to be pretty crazy for a console) whereas Microsoft has gone for a smart wide approach with many more compute units, with a more manageable/reasonable clock speed. Even the emphasis by Microsoft on VRS, yet another focus on being smart rather than just brute forcing it. VRS can save their GPU precious resources that can go back into improving performance.

Even looking at Microsoft's form factor, it's easy to say that's a sign of the kind of brute power they're going for, but then they also managed to make this thing as quiet as Xbox One X. It doesn't appear to me as if brute power and smart design are at odds with each other in the Xbox Series X design.

Even if Microsoft are building themselves a mini PC tower I still see efforts all around the design from what Microsoft has said thus far to keep from having to brute force their way to a result. There are efforts to take pressure off the hardware. I don't think they're taking that they have a lot of power for granted.

 

[DJ12]

I thought Sony was supposed to be ahead in devkit progression according to a lot of the insiders, this would contradict that point.

Pretty sure thats not been said (how are insiders supposed to know how many devkit revisions there are going to be or the exact performance of the current ones?), I've read the tools are more mature and easier to use, and devkits have probably been out longer with equivalent hardware in (the 2ghz APUs not the final retail ones - this is because Sony don't have direct x which means they can't just stick anything in the box of equivalent power like MS can), but pretty sure xbox are the only ones that are shipping with final silicon in at present. Could be wrong this information isn't really out there.

 

[sonomamashine]

Don't know how you could make that assumption baring in mind we know absolutely nothing about Sony's raytracing tech other than in December 2018 they stated they've been working on it for a while and were still working on it

Moving data around chip to chip requiers more energy = more power consumption for a console

Anyway, that's the concern of my electric company and the bank manager.

... and ? You suppose to sound smart ?
I think We are discussing two different things here, im saying that it gonna be very hot and power consuming for a console, on the other hand you discussing the fact that it wont matter to you if you are paying a few dollars more on your electricity bill.

 

[webber]

HDMI 2.1 is a given but good to see confirmation.
7:05 mark.

 

[demigod]

Or semi custom 36 dual cu Apu.

Please note unseen GPU with a test CPU pretty much confirms this is an APU.

Now which console makers chips get tested on windas and has vr. Hmm I wonder.....

Insane Metal kept hinting about the 4800 APU. I really can't think of which company would want VR in their next machine .

 

[Disco_]

Do you have a link with that doc?

docs were removed. I don't have a copy. Some on ree have saved it and shared some info after it was removed. Whatever data there was on arden/sparkman was theoretical data, not test data.

Or semi custom 36 dual cu Apu.

Please note unseen GPU with a test CPU pretty much confirms this is an APU.

Now which console makers chips get tested on windas and has vr. Hmm I wonder.....

It'a a laptop with an amd apu and a 2060 inside of it. If you google the eng sample (100-000000098-40_39/27_Y ) it'll take you to a post from about a month ago mentioning said laptop and everything inside it

ASUS Zephyrus G GA401IV DxDiag - Technopat Sosyal

------------------ System Information ------------------ Operating System: Windows 10 Pro 64-bit (10.0, Build 18363) (18362.19h1_release.190318-1202) Language: English (Regional Setting: English) System Manufacturer: ASUSTeK COMPUTER INC. System...

www.technopat.net

 

[ArabianPrynce]

Jason said no inside info for his prediction

 

[HeisenbergFX4]

Jason said no inside info for his prediction

The reason I found it interesting was because of this, no insider info that they are for sure attending which I would think would have been decided one way or another by now.

 

[joe_zazen]

.

The reason I found it interesting was because of this, no insider info that they are for sure attending which I would think would have been decided one way or another by now.

they just hired sales and marketing lead this week, she will need time to get up to speed and make decisions. I wouldn't expect anything for a month+.

 

[DonF]

looks like crap, maybe it looks better in motion...nahh

 

[DJ12]

Moving data around chip to chip requiers more energy = more power consumption for a console

... and ? You suppose to sound smart ?
I think We are discussing two different things here, im saying that it gonna be very hot and power consuming for a console, on the other hand you discussing the fact that it wont matter to you if you are paying a few dollars more on your electricity bill.

I was making a joke. you have absolutely no idea how much energy it uses or heat it produces, we don't know nothing about it. I can tell you when I turn RTX on my card uses less power because the frame rates are lower. With vsync enabled and set to 60 fps it uses more 20% more power (Monitoring software only shows percentage of available power used) though to be fair and is slightly hotter 46v51 degrees.

I used Shadow of the Tomb Raider for this test with everything set to ultra.

Insane Metal kept hinting about the 4800 APU. I really can't think of which company would want VR in their next machine .

The code points to the 4800H apu, but the clocks are lower, and we already know that chip is beyond the engineering sample stage. (could be an old test though)

I expect with all the leaks and dataminers, AMD are actively trying to hide what the chips are for now. Other than ruling out this being scarlet, as why would you test VR with that chip, I wouldn't be surprised if this was the PS5 chip. (I also wouldn't be surprised if it wasn't, I'm not invested in it)

One thing we do know though is with only 7 Radeon CUs, no way on earth is a 4800h destroying a 2080ti in any test.

 

[Disco_]

The code points to the 4800H apu, but the clocks are lower, and we already know that chip is beyond the engineering sample stage. (could be an old test though)

I expect with all the leaks and dataminers, AMD are actively trying to hide what the chips are for now. Other than ruling out this being scarlet, as why would you test VR with that chip, I wouldn't be surprised if this was the PS5 chip. (I also wouldn't be surprised if it wasn't, I'm not invested in it)

One thing we do know though is with only 7 Radeon CUs, no way on earth is a 4800h destroying a 2080ti in any test.

Already answered you above. It's a laptop with an APU and a 2060.

 

[DJ12]

docs were removed. I don't have a copy. Some on ree have saved it and shared some info after it was removed. Whatever data there was on arden/sparkman was theoretical data, not test data.




It'a a laptop with an amd apu and a 2060 inside of it. If you google the eng sample (100-000000098-40_39/27_Y ) it'll take you to a post from about a month ago mentioning said laptop and everything inside it

ASUS Zephyrus G GA401IV DxDiag - Technopat Sosyal

------------------ System Information ------------------ Operating System: Windows 10 Pro 64-bit (10.0, Build 18363) (18362.19h1_release.190318-1202) Language: English (Regional Setting: English) System Manufacturer: ASUSTeK COMPUTER INC. System...

www.technopat.net

Fair enough, still cannot see that setup beating a 2080ti, and it states the GPU used was AMD, which only uses GCN and 7 CUs.

It doesn't add up, but not invested in it to be fair so not going to be arguing the case any more than that.

Already answered you above. It's a laptop with an APU and a 2060.

Like I implied, doesn't scream 2080ti killer to me, even with both GPUs combined which of course wouldn't happen anyway.

 

[Disco_]

Fair enough, still cannot see that setup beating a 2080ti, and it states the GPU used was AMD, which only uses GCN and 7 CUs.

It doesn't add up, but not invested in it to be fair so not going to be arguing the case any more than that.

ASUS Zephyrus G GA401IV DxDiag - Technopat Sosyal

------------------ System Information ------------------ Operating System: Windows 10 Pro 64-bit (10.0, Build 18363) (18362.19h1_release.190318-1202) Language: English (Regional Setting: English) System Manufacturer: ASUSTeK COMPUTER INC. System...

www.technopat.net

Look at the link above. It lists everything in that laptop. The onboard APU and the nvidia gpu. Just scroll though the entire spec list.

I doesn't add up but it is what it is.