[Benchlife] R9 480 (Polaris 10 >100w), R9

[IllogicalGlory]

I think people are worried, and perhaps justifiably so, that developers are going to be putting more work into the Neo versions of their games than the PS4 classic versions. Many have enough trouble developing games for one/two consoles, I'm not optimistic about their chances for developing for two/three.

I think that either we'll see games mainly developed for the classic that don't really take advantage of the Neo (just higher resolution, more stable 30 FPS), or games developed mainly for the Neo where performance/visuals are poor on the classic. I'm not confident we'll see games that work well on the PS4 and really take advantages of the Neo's hardware at the same time. Perhaps we'll be seeing the big studios working on Neo versions of the game and outsourcing the current gen versions like we did back when the PS360 were still getting new games.

The hardware isn't going to change, but software is where the experience is made, or broken.

 

[Gonad the Barbarian]

Also do we say gtx 950/960 owners are second class citizens? In an age where games are ported to pc and handle multiple graphics cards, I'm sure they can add one more console configuration that's basically similar.

I mean if it was so hard more games would be ps4/xbone exclusives instead of cross platform

Ya'll keep saying this is OK by likening it to the PC experience. That's the problem.

 

[swilli89]

Also do we say gtx 950/960 owners are second class citizens? In an age where games are ported to pc and handle multiple graphics cards, I'm sure they can add one more console configuration that's basically similar.

I mean if it was so hard more games would be ps4/xbone exclusives instead of cross platform

Exactly!! Dev's can easily target two performance levels.. they already are actually: PS4 is 50% more shader power than XBONE already, this isn't a big deal at all for current owners.

I think people are worried, and perhaps justifiably so, that developers are going to be putting more work into the Neo versions of their games than the PS4 classic versions. Many have enough trouble developing games for one/two consoles, I'm not optimistic about their chances for developing for two/three.

I think that either we'll see games mainly developed for the classic that don't really take advantage of the Neo (just higher resolution, more stable 30 FPS), or games developed mainly for the Neo where performance/visuals are poor on the classic. I'm not confident we'll see games that work well on the PS4 and really take advantages of the Neo's hardware at the same time. Perhaps we'll be seeing the big studios working on Neo versions of the game and outsourcing the current gen versions like we did back when the PS360 were still getting new games.

The hardware isn't going to change, but software is where the experience is made, or broken.

This is mainly for VR and 4K. I would wager that most games simply being rendered at 4k instead of 1080P will be enough to tax the hardware.

 

[tential]

Yes exactly the ps4 has far more gpu power yet devs still manage it. I forgot to mention that thanks.

 

[96Firebird]

I don't think the "new" consoles will be rendering games at 4K...

 

[tential]

I don't think the "new" consoles will be rendering games at 4K...

Obviously not. Next gen maybe but not this gen. But it will give us vastly improved textures overall for gaming. Or so one prays....

 

[maddie]

I think people are worried, and perhaps justifiably so, that developers are going to be putting more work into the Neo versions of their games than the PS4 classic versions. Many have enough trouble developing games for one/two consoles, I'm not optimistic about their chances for developing for two/three.

I think that either we'll see games mainly developed for the classic that don't really take advantage of the Neo (just higher resolution, more stable 30 FPS), or games developed mainly for the Neo where performance/visuals are poor on the classic. I'm not confident we'll see games that work well on the PS4 and really take advantages of the Neo's hardware at the same time. Perhaps we'll be seeing the big studios working on Neo versions of the game and outsourcing the current gen versions like we did back when the PS360 were still getting new games.

The hardware isn't going to change, but software is where the experience is made, or broken.

This is probably the main reason. Based on past consoles, it is understandable. Sony needs to explain in the simplest terms exactly what compatibility means for the X86 based consoles and how different it is from previous versions.

Having the CPU very close to the original version means that all the AI, physics, etc is the same between console versions. The graphical output is where the main difference between versions will be observed.

Sony's PR dept had better get to work.

 

[swilli89]

I don't think the "new" consoles will be rendering games at 4K...

And why not? If PS4 Neo includes a Polaris 10 class GPU which was shown to be comparable to a Fury X then it can definitely do some games in 4k. The main point here is that comparison is not optimized. We are discussing Polaris 10 at Fury X speeds in a desktop windows environment. With much closer to metal programming and impending DX 12 performance tweaks this new chipset could absolutely run games @ 4k with medium-ish settings.

This is probably the main reason. Based on past consoles, it is understandable. Sony needs to explain in the simplest terms exactly what compatibility means for the X86 based consoles and how different it is from previous versions.

Having the CPU very close to the original version means that all the AI, physics, etc is the same between console versions. The graphical output is where the main difference between versions will be observed.

Sony's PR dept had better get to work.

Yes smart move to just keep the cpu but tweaked up 30% clockspeed. With console optimizations this should continue to be sufficient in most applications.

 

[HurleyBird]

I don't think the "new" consoles will be rendering games at 4K...

Even if they could, they shouldn't be unless you're looking at simpler games that aren't trying to push the graphical envelope. There are serious diminishing returns from pumping up the resolution that high on a TV. You're better off targeting ~1500p, upscaling, and spending your performance budget on more meaningful things.

 

[monstercameron]

Even if they could, they shouldn't be unless you're looking at simpler games that aren't trying to push the graphical envelope. There are serious diminishing returns from pumping up the resolution that high on a TV. You're better off targeting ~1500p, upscaling, and spending your performance budget on more meaningful things.

not everything has to be rendered in 4k...things like the hud and on screen text could take advantage of the added sharpness.

 

[dark zero]

Propably this.

Yeah, that is.

 

[Madpacket]

I'm not sure why this is complicated to understand.

VR and 4K are the new target markets.

PS4 is too weak to manage 4K and HDMI only supports 30hz. VR requires hacks (frame doubling) to handle VR which is not ideal.

This 4.5 is simply a more powerful version of the PS4 to handle 4K @ 60Hz and proper VR.

What else would it be for?

As for compatibility this is very easy or Sony. Same libraries, engines, middleware will all work on a more powerful X86 CPU with updated GCN graphics. If VR or 4K gaming takes off then this will be a hit. If 4K or VR fail then it won't sell.

 

[sm625]

I still dont believe the 14nm PS4K. 36 CU @ 911MHz fits well within the thermal envelope of 28nm. Consider the fact that R9 Nano is 64 CU @ 1GHz @ 175W (meaning the actual GPU portion is only ~150W). That means 36 CU @ 900MHz would easily fit within a 100W envelope, which is well within the realm of realistic power consumption for a console. That leaves 50W for the other components to keep it the same as the current PS4. Remember that 28nm has improved dramatically since PS4 SoC first fabbed. They will have no problems with yields at 900MHz. There is absolutely no reason to expect PS4K to be on 14nm, it goes against all past history of how they design these things.

 

[HurleyBird]

not everything has to be rendered in 4k...things like the hud and on screen text could take advantage of the added sharpness.

Well, yeah, that goes without saying

 

[MrTeal]

I still dont believe the 14nm PS4K. 36 CU @ 911MHz fits well within the thermal envelope of 28nm. Consider the fact that R9 Nano is 64 CU @ 1GHz @ 175W (meaning the actual GPU portion is only ~150W). That means 36 CU @ 900MHz would easily fit within a 100W envelope, which is well within the realm of realistic power consumption for a console. That leaves 50W for the other components to keep it the same as the current PS4. Remember that 28nm has improved dramatically since PS4 SoC first fabbed. They will have no problems with yields at 900MHz. There is absolutely no reason to expect PS4K to be on 14nm, it goes against all past history of how they design these things.

The Nano also has the benefit not running GDDR5 or a GDDR5 controller, and the CUs are often underutilized as they're bottlenecked by the front end. A PS4 pulls 130-140W at the wall while gaming, so doubling the CU count at 28nm while staying at the same power consumption seems unlikely.

 

[Adored]

Power and clocks are well within 28nm ranges while the whole unit could still stay below 200W. Jaguar has been 2.05GHz for 2 years now.

The issue is the die size, which could be around 450mm2 on 28nm. 14nm should be an awful lot smaller (I read 14FF+ is around 30% the size of Globalfoundries 28nm) so it's likely worth it even with the relatively newer status of the node.

 

[3DVagabond]

No mention on he article, but I tink that Freesync would be a huge plus in this revision.

Might explain AMD adding the feature to HDMI?

 

[swilli89]

Power and clocks are well within 28nm ranges while the whole unit could still stay below 200W. Jaguar has been 2.05GHz for 2 years now.

The issue is the die size, which could be around 450mm2 on 28nm. 14nm should be an awful lot smaller (I read 14FF+ is around 30% the size of Globalfoundries 28nm) so it's likely worth it even with the relatively newer status of the node.

The size alone of this APU is the limiting factor. I doubt AMD would offer this upgrade path as it would probably maintain the razor thin margins they are already accepting. I would think it would make sense to at least port the design over to 14nm or 16nm given the assumed high volumes of a playstation console. The timing is just too hard to ignore, it just makes no sense to release on 28nm.

 

[AtenRa]

PS4K Could be 28nm HDL.

 

[geoxile]

PS4K Could be 28nm HDL.

Pretty sure GPUs already use HDL or even denser layout.

 

[MrTeal]

PS4K Could be 28nm HDL.

Given that the PS4 APU is essentially a big GPU with some small Cat cores added in, is there a reason to think that it isn't already using HDL libraries?

 

[swilli89]

Given that the PS4 APU is essentially a big GPU with some small Cat cores added in, is there a reason to think that it isn't already using HDL libraries?

Yes and it even shows that in the slide. Its saying "we are using the same library density techniques that typically we use in GPUs". I think we are looking at a 14nm PS4 refresh.

 

[AtenRa]

Really i have no idea, i speculate since the PS4 and XB1 SoCs also have 8x CPU Cores they dont use HDL. I believe the first APU that used HDL was the Carrizo and not the Console APUs.

Edit: Personally i would say it depends on when the new Consoles are due for release. If they are 2016 i would say they are 28nm, if they are to be released in late 2017 or early 2018 they probably are going to be 14/16nm FF.

 

[ShintaiDK]

And why not? If PS4 Neo includes a Polaris 10 class GPU which was shown to be comparable to a Fury X then it can definitely do some games in 4k. The main point here is that comparison is not optimized. We are discussing Polaris 10 at Fury X speeds in a desktop windows environment. With much closer to metal programming and impending DX 12 performance tweaks this new chipset could absolutely run games @ 4k with medium-ish settings.

Your expectations of Polaris seems to have reached new heights.

The PS4 today already struggles at 1080p. Adding twice the power wont do anything if you are to render a resolution 4 times greater.

 

[Headfoot]

Doesn't make a lot of sense to backport Polaris to 28nm when its 14nm native, and the economics of the consoles already dictate going to 14nm for #of die per wafer and power consumption reasons. I doubt it will be a 28nm, considering how touchy they are about thermal budgets after the PS360 generation fiascos