• 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.
Status
Not open for further replies.

Deleted member 1589

User requested account closure
Banned
Oct 25, 2017
8,576
Just realised that's probably how Sony will handle emulation from now on. It's a rather smart idea for the time being.
 

sncvsrtoip

Banned
Apr 18, 2019
2,773
A flop is a flop. If you took an Xbox one x and an Xbox series x and ran them through a benchmark the x would achieve a 6TF score where as the xsx would achieve 12TF.
What, what benchmark ?:d Benchmarks usually mesaure frames per seconds and gives score according to numbers of frames so for example in benchmarks 9.5tf navi would have 2x frames number vs 6tf gcn. TFlops are therorectical number you calculate on paper (nvidia on paper calculate 1.8tf for rsx ;d)
 

DukeBlueBall

Banned
Oct 27, 2017
9,059
Seattle, WA
A flop is a flop. If you took an Xbox one x and an Xbox series x and ran them through a benchmark the x would achieve a 6TF score where as the xsx would achieve 12TF. Improvements in chipset architecture are accounted for in order to get to that 12TF rating. Twisting chipset improvements to mess with performance ratings is some needless fuckery

+1

I'm a random idiot on the Internet with no insider information to speak of.

Straight Outta Nowhere
 

Bosch

Banned
May 15, 2019
3,680
A flop is a flop. If you took an Xbox one x and an Xbox series x and ran them through a benchmark the x would achieve a 6TF score where as the xsx would achieve 12TF. Improvements in chipset architecture are accounted for in order to get to that 12TF rating. Twisting chipset improvements to mess with performance ratings is some needless fuckery
This doesn't make any sense. Benchs measure FPS not flops.

a 9 tflops RDNA will perform exactly as a 12 tflops GCN because of efficiency.

DF showed that on their video with a 5700XT they can double the performance of a Xbox ONE X
 

bcatwilly

Member
Oct 27, 2017
2,483
If that chip information is accurate and relevant to PS5 then it does raise some question as to whether some of the original Sony "target specs" information being shared was actually a GCN TF number instead of RDNA originally which would make it higher than the chip information and/or some of the comparisons from dev kits are based around what was a Vega 12-13 GCN TF based PS5 dev kit for example.
 

mindcontrol

Alt-account
Banned
Dec 17, 2019
11
Did anyone at least think that what rogame found is OUT OF DATE and just to make some people happy because XSX is more powerful according to that possibly OUT OF DATE info?
 

VX1

Member
Oct 28, 2017
6,999
Europe
Imo 12 Tflops of Xbox Series X is RDNA. They did not give an exact number because they aren't sure it will be 11.8/11.9 or 12.1 for example. PS3 and Xbox 360 did not reach the exact target better wait to be sure.

For PS5, I believe at lease jschreier and it is at least 10.8 Tflops. And Ariel, Ariel B0 and Oberon are probably devkit value.
Did Sony ever have higher specs in retail console comparing to dev kits? Or it's mostly vice versa?

End of year predictions:

Following the blueprint of the PS4's success, Sony went into this wanting to make a box for $400 and not lose money on it. I think 9-10TF is likely and BC with PS4/Pro will be 100% but offer no significant benefits due to the method used (CPU downclocks). People will figure out a way to download P.T. again similar to the old/current method and so there will be no merit in keeping your old hardware. It will sell 100-120m units being consistently strong across all regions. Any power deficit compared to the XSX won't matter one jot to the average consumer and with some of Sony's first party games you'd be forgiven for thinking that the power deficit went the other way.

Having been humiliated this generation Microsoft were keen to have the most powerful hardware and going forward it will be part of their core identity. XSX will be $500 and Lockhart will be $300, the latter being by far the more popular iteration. The efforts of the BC team will continue to amaze with a 9x res bump across the board for all 360 games and smoothed out frame rates and maxed out variable resolutions for Xbox One games. It will do similar numbers to the 360, clawing back marketshare in English speaking countries while struggling everywhere else and sell around 80m units. Given that Microsoft also have PC and ambitions for Xcloud and Game Pass that console userbase number means less than it did in the past. Also those who care about the power difference will be those who spend a lot of money on the platform so they'll feel justified in targeting that demographic and getting them entrenched in their ecosystem going forward.

Overall both platform holders will be very happy and will make bucketloads of money.
Sounds about right.
 

Winstano

Editor-in-chief at nextgenbase.com
Verified
Oct 28, 2017
1,828
If you put a 500bhp engine in an 18 wheeler, it will hardly move it. If you put a 500bhp engine in a lightweight car, it'll move like shit off a shovel.

That's the comparison people should be making - a flop is a fixed unit. It's how they're used that makes the difference. Make a more efficient architecture, get more out of it.
 

anexanhume

Member
Oct 25, 2017
12,912
Maryland
oh I'm aware. Btw need your input on the Power VR Rt solution to Oberon.

You should be asking chris 1515 in my opinion. My novice take is that they have a complete pipeline for RT thanks to the Caustic legacy. They've been making hardware RT for almost a decade at this point, and John Carmack thought their tech half a decade ago was solid.

anexanhume
I saw your comments about "silicon not final".
This leads up to some questions for me:
  1. Can you provide an assumption or explanation what you think will change for the final design?
  2. And why would you produce incomplete silicon for testing and validation? (referencing missing RT and VRS)
I would like to learn about the reasoning behind your thoughts there.

I think there are two possibilities. One is that they always conceived of pushing out a non-final APU for developers to play with. This could be an off the shelf Zen 2, IO, and Navi dies, with potentially only changes to the IO die needed. Additionally, it could be pin compatible with the final product. This would enable PS5-like silicon in their hands up to a year before they get the real development chips.

As a thought experiment - why does Navi 12 exist? A Navi 10 like design with hardware updates and HBM memory. Who is the desktop consumer for a product with high memory cost? Unless the memory cost isn't actually that high anymore, the product is simply a leveraged version of something used elsewhere, or both.

The second possibility is that this is plan B. At some point in early 2018 (At the latest), they called an audible and decided that Navi 10 wasn't good enough. Could this be a reaction to what they thought MS was cooking, how the narrative of X vs. Pro was evolving, developer feedback (this would have been the time Cerny was touring developers), or AMD pulling some roadmap items in as they went through some tumultuous times? Who knows. Small snippets like Chinese forum posters alleging a hardware revision would cut it close for 2020 launch are potential clues, though I am hesitant to bite on those things too hard.

What I do know is that virtually every insider is saying PS5 and XSX are close, and most of those are sticking to that story after the XSX reveal. We know hardware RT is confirmed. Sony said it, Matt said it, Kleegamefan said it. One of our grandmothers probably said it. Klee has also said VRS is included, and everyone is suggesting >10TF. All of these things Navi 10 is not. It would be surprising to see this many people with other valid info turn up wrong on this.

And to make my lack of allegiances clear - I am team parity. I own Pro and X and love them both.
 

Deleted member 2379

User requested account closure
Banned
Oct 25, 2017
1,739
If that chip information is accurate and relevant to PS5 then it does raise some question as to whether some of the original Sony "target specs" information being shared was actually a GCN TF number instead of RDNA originally which would make it higher than the chip information and/or some of the comparisons from dev kits are based around what was a Vega 12-13 GCN TF based PS5 dev kit for example.

This paste has been up for a bit and is where I believe the Original leaked 13TF numbers came from.

https://pastebin.com/w3fwc7Cb
 

gundamkyoukai

Member
Oct 25, 2017
21,069
Doesn't have to be in the APU. Can be a separate chip

That make no fucking sense.
Sony has been working with AMD for years and would have there road map in advance which is why the add stuff to PS4 and PS4 pro.
But decide they going to spend money and do RT and VRS on a separate chip because they like spending extra money and making there board more complex.
Mean while MS has everything from AMD but Sony decide nope lets do own thing which also make thing more complex for devs which Cerny does not like.
 
Last edited:

Deleted member 1589

User requested account closure
Banned
Oct 25, 2017
8,576
So the gist of it...

Playstation 4 emulation relies on fooling the system into treating CUs as workgroups. So if there's a midgen refresh for example, they can theoretically make it work by keeping to the butterfly CU pattern, and as long as it has the same amount of workgroups as the PS4 Pro, then it could work.

So basically, you'd have 20 workgroups with 3 CUs inside, and disable 2 workgroups for yields. Leaves you with 54 CUs in 18 WGP, thus 9 WGP in each side.

The problem is I'm not sure if there's such a thing as a 60 CU GPU or if it's even possible.
 
Aug 26, 2019
6,342
Having been humiliated this generation Microsoft were keen to have the most powerful hardware and going forward it will be part of their core identity.
Yep this is why I think MS will have the power lead next-gen. They were embarrassed that a launch game like Battlefield 4 had to run at sub-1080p resolution on their console and wanted to prevent a similar situation from ever happening again.
 

Deleted member 12635

User requested account closure
Banned
Oct 27, 2017
6,198
Germany
You should be asking chris 1515 in my opinion. My novice take is that they have a complete pipeline for RT thanks to the Caustic legacy. They've been making hardware RT for almost a decade at this point, and John Carmack thought their tech half a decade ago was solid.



I think there are two possibilities. One is that they always conceived of pushing out a non-final APU for developers to play with. This could be an off the shelf Zen 2, IO, and Navi dies, with potentially only changes to the IO die needed. Additionally, it could be pin compatible with the final product. This would enable PS5-like silicon in their hands up to a year before they get the real development chips.

The second possibility is that this is plan B. At some point in early 2018 (At the latest), they called an audible and decided that Navi 10 wasn't good enough. Could this be a reaction to what they thought MS was cooking, how the narrative of X vs. Pro was evolving, developer feedback (this would have been the time Cerny was touring developers), or AMD pulling some roadmap items in as they went through some tumultuous times? Who knows. Small snippets like Chinese forum posters alleging a hardware revision would cut it close for 2020 launch are potential clues, though I am hesitant to bite on those things too hard.

What I do know is that virtually every insider is saying PS5 and XSX are close, and most of those are sticking to that story after the XSX reveal. We know hardware RT is confirmed. Sony said it, Matt said it, Kleegamefan said it. One of our grandmothers probably said it. Klee has also said VRS is included, and everyone is suggesting >10TF. All of these things Navi 10 is not. It would be surprising to see this many people with other valid info turn up wrong on this.

And to make my lack of allegiances clear - I am team parity. I own Pro and X and love them both.
Thanks for coming back to me and taking the time to explain where you come from.

I may come back to it in the future.

hope you enjoyed the holidays so far.
 
Last edited:

chris 1515

Member
Oct 27, 2017
7,074
Barcelona Spain
Did Sony ever have higher specs in retail console comparing to dev kits? Or it's mostly vice versa?


Sounds about right.

Normally they have the same with final devkit but before the nearly final devkit, it was often different.

For example, on PS4 the first SOC based devkit was available in January 2013 but the frequency and the RAM was not the same.

vgleaks.com

ORBIS Devkits Roadmap/Types

Sony and Microsoft have had a lot of work with Orbis and Durango. Create a new hardware it’s a long running process, you have to build different versions of hardware and make changes along the core until the last version, the closer one to the retail product. We have created an article with...

We will know better in January probably. Here it means what we have in 2019 is probably not exactly like the final architecture.

No Sony dev kit specs are normally the same it was that way for PS4.

Not all devkit, it depends of the devkit. Before January 2013, the PS4 devkit was different than the PS4. It was some PC part.
 
Last edited:
Dec 8, 2018
1,911
There are a bit to many things about this info that does not make sense to me. Like that the devkit that is supposed to be the PS5 is literally running as fast as it possibly can with a clock of 2Ghz which would mean low yields and a very big/effective cooling solution, that the hardware specs have no mention of ray tracing capabilities or VRS when the other (Arden) do when confirmed by Cerny that it will have at least ray tracing, how even known insiders that basically never are wrong and don't typically report or comment on "rumors" (Jason) have heard from developers that they would be extremely similar and now we have a 30% tflop difference including missing ray tracing capabilities in the hardware and VRS.

If this is actually the PS5 dev kit and I would have to put money on something it would be that Sony wanted to provide developers with something to use as early as possible and ran with the best AMD card available at the time clocked as high as possible to bring the performance "close" to target specs without developing a completely new chip for developers before even AMD had their own "big" one and perhaps have some hardware outside the APU to provide the ray tracing capabilities while waiting for AMD chip to mature.

I certainly don't believe he's manufacturing these specs and found them but I just can't wrap them around basically everything else we heard from multiple directions and how it would fit what we have been told from people in the industry for the last 6 months.
 

disco_potato

Member
Nov 16, 2017
3,145
$299 @ 4TF
$399 @ 10TF
$499 @ 12TF
Plus beefy CPU and SSD in each form.

How does that fly for everyone?

I'll have that PS5 PRO in 2024 please.

Hey guys,

How do you think the audio 3D will work ? Is it hardware implementation so you will need a specific headset made for/by Sony or is it just a virtual sound like DTS surround and other 7.1 virtualized ?

I would like to grab a XM3 or something else Bluetooth headsets with noise canceling. But I'm not sure if Ps5 will finally accept all Bluetooth hardware ? Shame that PS4 can't be used with the Sony XM3... with a dongle maybe ...?

As long as aptX is supported by the transceiver/dongle it should work.
 

VX1

Member
Oct 28, 2017
6,999
Europe
Thanks for coming back to me and taking the time to explainwhere you come from.

I may come back to it in the future.

hope you enjoyed the holidays so far.
anexanhume has a very interesting hypothesis.It would be fascinating if it turns out to be true but i am, frankly, deeply skeptical.We have so little time left till mass production starts.Anyway,fun times ahead! :)
 

Deleted member 2379

User requested account closure
Banned
Oct 25, 2017
1,739
So the gist of it...

Playstation 4 emulation relies on fooling the system into treating CUs as workgroups. So if there's a midgen refresh for example, they can theoretically make it work by keeping to the butterfly CU pattern, and as long as it has the same amount of workgroups as the PS4 Pro, then it could work.

So basically, you'd have 20 workgroups with 3 CUs inside, and disable 2 workgroups for yields. Leaves you with 54 CUs in 18 WGP, thus 9 WGP in each side.

The problem is I'm not sure if there's such a thing as a 60 CU GPU or if it's even possible.

I posted this awhile ago

http://www.redgamingtech.com/playstation-5-gpu-specs-leak-2ghz-backwards-compatibility-mode/
 

anexanhume

Member
Oct 25, 2017
12,912
Maryland
anexanhume has a very interesting hypothesis.It would be fascinating if it turns out to be true but i am, frankly, deeply skeptical.We have so little time left till mass production starts.Anyway,fun times ahead! :)
Let me be clear I don't necessarily think that this is what has happened. This is me working backward from what insiders are telling us. Absent their info, I would believe PS5 is 36CU, 2.0GHz, with some type of hardware RT, but no VRS, and less than 10TF.
 

Deleted member 2379

User requested account closure
Banned
Oct 25, 2017
1,739
There are a bit to many things about this info that does not make sense to me. Like that the devkit that is supposed to be the PS5 is literally running as fast as it possibly can with a clock of 2Ghz which would mean low yields and a very big/effective cooling solution, that the hardware specs have no mention of ray tracing capabilities or VRS when the other (Arden) do when confirmed by Cerny that it will have at least ray tracing, how even known insiders that basically never are wrong and don't typically report or comment on "rumors" (Jason) have heard from developers that they would be extremely similar and now we have a 30% tflop difference including missing ray tracing capabilities in the hardware and VRS.

If this is actually the PS5 dev kit and I would have to put money on something it would be that Sony wanted to provide developers with something to use as early as possible and ran with the best AMD card available at the time clocked as high as possible to bring the performance "close" to target specs without developing a completely new chip for developers before even AMD had their own "big" one and perhaps have some hardware outside the APU to provide the ray tracing capabilities while waiting for AMD chip to mature.

I certainly don't believe he's manufacturing these specs and found them but I just can't wrap them around basically everything else we heard from multiple directions and how it would fit what we have been told from people in the industry for the last 6 months.
Take one look at the dev kit leak. It's built around a massive cooling solution.
 

VX1

Member
Oct 28, 2017
6,999
Europe
Let me be clear I don't necessarily think that this is what has happened. This is me working backward from what insiders are telling us. Absent their info, I would believe PS5 is 36CU, 2.0GHz, with some type of hardware RT, but no VRS, and less than 10TF.
You have spoken ;)

Btw,i am still fascinated they might really go with 2GHz clock in a retail console...
 

gundamkyoukai

Member
Oct 25, 2017
21,069
Not all devkit, it depends of the devkit. Before January 2013, the PS4 devkit was different than the PS4. It was some PC part.

Was talking about the final dev kits .
Since the devs kits are given way in advance and things change from the early version .( as we saw even with PS5)
But final dev kits have the same specs as the system for at least PS4.
 

Simon-chan

Member
Oct 27, 2017
2,044
Italy
QaGILcz.gif
 

chris 1515

Member
Oct 27, 2017
7,074
Barcelona Spain
Was talking about the final dev kits .
Since the devs kits are given way in advance and things change from the early version .( as we saw even with PS5)
But final dev kits have the same specs as the system for at least PS4.

Yes exactly the same but we speak of May/June 2020 for final devkit. But in January we must have a devkits looking like what will be inside the PS5 before they can work with PC part.
 

sncvsrtoip

Banned
Apr 18, 2019
2,773
Let me be clear I don't necessarily think that this is what has happened. This is me working backward from what insiders are telling us. Absent their info, I would believe PS5 is 36CU, 2.0GHz, with some type of hardware RT, but no VRS, and less than 10TF.
since when ?:d I remember you claim that gonzalo is not navi10lite
 
Status
Not open for further replies.