PS4 Pro Speculation (PS4K NEO Kaio-Ken-Kutaragi-Kaz Neo-san)

Status
Not open for further replies.
A die shrink is still a die shrink, all the "benefits" of die shrinks still apply.
...
In fact going off topic and against the corporate brass decisions. AMD should have die shrunk both their Phenom II X4 and X6 CPUs to 32nm (down from 45nm)

AMD did shrink K10 to 32nm for Llano. They had yield issues on the CPU side and clock speeds were disappointing, rendering them uncompetitive.

These days, die shrinks aren't necessarily a magic wand that makes everything massively faster and cooler. 32nm to 28 nm at GF made design choices that prioritised density increases (suiting GPU) but hurt frequency at the top of the CPU spectrum.

The top end Jaguar on desktop hits 2.05 gHz, and that sees a bump for the APU power to 25W from 15W. The top end Puma is 2.2 gHz, and that's desktop only with a 25W TDP for the APU. Sony aren't doing badly with 2.1 gHz for 2 Jaguar modules - they're likely at the point where either power would start to sky-rocket or yields would drop off. Jaguar simply isn't designed for high clocks, and Sony have a sizeable GPU in there needing power too.
 
Jaguar was never intended as a high frequency part because it was never supposed to be a high wattage part. It's quite possible it won't go much higher than 2.1GHz without lengthening the pipeline. Sony aren't going to go messing with the cores now.
Jaguar can safely go to 2.4ghz.
 
...Or putting it another way, is it really a better experience to have someone invent an awesome technique that enables their end-of-life 20 fps game to reach 24 fps over having better hardware for that game that runs it at 60 fps with the traditional techniques? From an intellectual POV, the clever software solutions are far more exciting, but from an experience POV the better game experience is more worthwhile.
....
If COD manages 30 fps on PS4 and a new MOH plays just as well but at 60 fps using better tech (and higher res, better quality on Neo), MOH could steal COD players. So competition is still going to be a mother of invention even if necessity isn't.

Agree with everything you say. I was genuinely counting on clever developers like Sebbbi to increase longevity of my purchase. Now this won't be the case, I'll have to shell out money for better hardware if software side of things stagnate. We're going to get better looking games either way, but I'll be left with less money in my pocket that I could use to buy games. Hooray for competition between devs though :) But I don't think it would be as big driving force as trying to cram everything to a fixed spec. API's getting fatter is indeed a problem though - forward compatibility has the potential to really pull things back: This may mean games on PS5 never utilizing, for example, convservative rasterization if the developer does not want to maintain multiple render paths, even if that feature is added to PS5 hardware.
 
Last edited:
AMD did shrink K10 to 32nm for Llano. They had yield issues on the CPU side and clock speeds were disappointing, rendering them uncompetitive.

These days, die shrinks aren't necessarily a magic wand that makes everything massively faster and cooler. 32nm to 28 nm at GF made design choices that prioritised density increases (suiting GPU) but hurt frequency at the top of the CPU spectrum.

The top end Jaguar on desktop hits 2.05 gHz, and that sees a bump for the APU power to 25W from 15W. The top end Puma is 2.2 gHz, and that's desktop only with a 25W TDP for the APU. Sony aren't doing badly with 2.1 gHz for 2 Jaguar modules - they're likely at the point where either power would start to sky-rocket or yields would drop off. Jaguar simply isn't designed for high clocks, and Sony have a sizeable GPU in there needing power too.
Firstly your opinions on 28nm only apply to carrizo as amd did work with glofo to customize their 28nm process for density and speed. Compared to carrizo that uses a more stock 28nm process. You can compare carrizo max sane clock of 3.5ghz compared to kaveris of 4.1ghz.

Also a note on your assessment of jaguar/puma. You are comparing tdps and not actual power draw. Tdp is a package deal not cpu cores only. It does use more power though.
 
The top end Jaguar on desktop hits 2.05 gHz, and that sees a bump for the APU power to 25W from 15W. The top end Puma is 2.2 gHz, and that's desktop only with a 25W TDP for the APU. Sony aren't doing badly with 2.1 gHz for 2 Jaguar modules - they're likely at the point where either power would start to sky-rocket or yields would drop off. Jaguar simply isn't designed for high clocks, and Sony have a sizeable GPU in there needing power too.
In addition to that, those 2.2Ghz 28nm Puma+ cores are using binning to get there. Dies with poorer frequency potential across all 4 cores end up as lower-tier SKUs, or as fused-off 2-core varients.

Sony can't bin like that, they are only after a single configuration and aren't fusing any of the CPU cores off. They need all 8 cores on 14nm to reach the target frequency and selling into a price-conscious market they can't afford to throw many dies away. All this means they need to be conservative with frequency.
 
Agree with everything you say. I was genuinely counting on clever developers like Sebbbi to increase longevity of my purchase. Now this won't be the case, I'll have to shell out money for better hardware if software side of things stagnate. We're going to get better looking games either way, but I'll be left with less money in my pocket that I could use to buy games. Hooray for competition between devs though :) But I don't think it would be as big driving force as trying to cram everything to a fixed spec. API's getting fatter is indeed a problem though - forward compatibility has the potential to really pull things back: This may mean games on PS5 never utilizing, for example, convservative rasterization if the developer does not want to maintain multiple render paths, even if that feature is added to PS5 hardware.

assuption again, they get games running better on XBO with half the market value return the PS4 will offer so why not bother again? Lazy devs?
 
Doubts are very justified at this point.

Remember that Sony needs to incentivise people to buy the new platform, while at the same time maintaining the original PS4 userbase.

If a game is 'too much better' on PS4K, and the PS4 version runs like crap, people might buy into the new system but I would expect the backlash against Sony to be similar to what MS experienced at the beginning to this generation.

If a game runs great on PS4, without much of a difference from the Neo version, then what's the point of buying the bloody thing?

They will need to strike the right balance, right in the middle of these two outcomes, and I think it will be a tough job.
 
assuption again, they get games running better on XBO with half the market value return the PS4 will offer so why not bother again? Lazy devs?
We can't do anything better than assuming things. This is all new territory. And all this makes is make Xbox One a bit more obsolete and force MS an Xbox 1.667, or maybe they'll hang their towel and make an Azure Streaming Box (of chocolate.)
 
Jaguar can safely go to 2.4ghz.

That's Puma, on turbo-ing parts, that are most likely carefully binned, that are operating at the extreme worst part of the perf/watt curve that the cores were ever forced to operate in.

What speed a binned, top end part can turbo to when much of the APU is disabled has no bearing on what speed is realistic for Sony to clock their two Jaguar modules to as part of a monster APU (more complex that anyone else has ever made).

"Safely" is a pretty meaningless descriptor and doesn't take into account any of the things we're actually discussing.

Firstly your opinions on 28nm only apply to carrizo as amd did work with glofo to customize their 28nm process for density and speed. Compared to carrizo that uses a more stock 28nm process. You can compare carrizo max sane clock of 3.5ghz compared to kaveris of 4.1ghz.

Also a note on your assessment of jaguar/puma. You are comparing tdps and not actual power draw. Tdp is a package deal not cpu cores only. It does use more power though.

You seem to be confusing a number of things. GF 28nm node is designed for density over the more clock focused 32nm. That's one of the reasons the 8 core, none APU, speed racer chips stayed on 32nm.

In addition to that, Carrizo itself features a large amount of engineering effort to improve density, power, turbo capability while targeting a lower power envelope than Kaveri. It's these decisions that in AMDs own words impacted on frequency scaling at the top end. It's the same 28nm node though.

I clearly sated that those were APU TDPs, so it's a little bizarre you're trying to bring that point up. The important part to note is that TDP increases disproportionately with clock speed. Performance per watt drops off badly as you try and hit those high clocks on Jaguar.

In addition to that, those 2.2Ghz 28nm Puma+ cores are using binning to get there. Dies with poorer frequency potential across all 4 cores end up as lower-tier SKUs, or as fused-off 2-core varients.

Sony can't bin like that, they are only after a single configuration and aren't fusing any of the CPU cores off. They need all 8 cores on 14nm to reach the target frequency and selling into a price-conscious market they can't afford to throw many dies away. All this means they need to be conservative with frequency.

Indeed. They need all cores to operate at a sane power level and with a single bin, and they can't rely on being "turboed" up to the kind clock speeds that would require the GPU to throttle back.
 
@hesido Sorry, the way you worded it could have been better than making it sound like a statement of fact.

@London-boy - 40 MILLION customers - that figure will only get bigger as the price drops further, the PS4k will only really be for folks with cash to splash and those who must have the latest and greatest (like GAF & here) to begin with. Also I'm sure the PS4 will be making a bigger margin for Sony.
 
The way you worded it could have been better then rather than making it like a statement of fact.

@London-boy - 40 MILLION customers - that figure will only get bigger as the price drops further, the PS4k will only really be for folks with cash to splash and those who must have the latest and greatest (like GAF & here)
I'm aware of that, and that's my point. They really need to strike the right balance if they want both platforms to succeed at the same time, without alienating a whole lot of customers.
 
@London-boy ok, I missread you...we were talking about devs not pushing PS4 and I was saying there would be 40 million customers so it's well worth them pushing PS4. As for Sony (which is what you were talking about which I missread as devs), like I said, this will be marketed as an option only and it seems they are being strict with devs on QC - what I will say is that it will be expected PS4 games still run better than XBO games (in general)...so we have a marker to compare against
 
@hesido Sorry, the way you worded it could have been better than making it sound like a statement of fact.
If it's not a fact that PS4 Neo will run games better, I take back everything that I said. You seem to have missed my point. (But yeah, I may need to word those better if you missed my point :p )
 
If it's not a fact that PS4 Neo will run games better, I take back everything that I said. You seem to have missed my point. (But yeah, I may need to word those better if you missed my point :p )

I'm sorry but I don't see how else to take this statement;

"I was genuinely counting on clever developers like Sebbbi to increase longevity of my purchase. Now this won't be the case, I'll have to shell out money for better hardware"

Of course Neo will run games better it has to! The point I'm making is devs will still be driving longevity because they have customers there (40M and ever rising) so it's in their interest.
 
If I was Sony, I would release it at the same price point as the current PS4 and just stop making the PS4.

I imagine AMD have made Sony a very attractive offer on the hardware to even make this a possibility so why not. No one is going to buy the less powerful PS4 anyway if they have a choice.
 
I'm sorry but I don't see how else to take this statement;

"I was genuinely counting on clever developers like Sebbbi to increase longevity of my purchase. Now this won't be the case, I'll have to shell out money for better hardware"

Of course Neo will run games better it has to! The point I'm making is devs will still be driving longevity because they have customers there (40M and ever rising) so it's in their interest.
I'm pessimistic and I have reasons to be, looking at the current game development scene. When I say "now this won't be the case", I think it should be obvious that this my personal assumption and belief because how the heck would I know how it will turn out, you have to deduce that this is my pessimistic forecast, but you talked as if I'm trying to pass my own assumptions as fact. Of course I'm not trying to word these as fact.

Unless MS are dropping the X1 this year, I'd say clever developers like sebbbi are going to be kept busy. :yep2:
Hopefully, for this gen, yes, until Console makers all go into a technology race in iterative fashion. Then of course development focus could shift (could :p ) from achieving efficiency on a fixed hardware, to more generic optimizations in graphics and whatnot, hindered by the forward compatibility requirements and not being able to use the latest hardware features because of the existence of the PS4. At least consoles used to set a bar for features with every major leap. Now this may not be (may :p ) case any more.
 
Then they'd lose a lower price of entry for their console. At this point in the platform's life, some really low prices would explode sales, and the tail is likely shorter than last gen. Well, we're tlaking upgraded console iterations now, so the tail wil definitely be shorter. Sony can't survive on just a high/mid tier pricing can they? Maybe they'd rely on only releasing high end and having second hand sales cover low end, but that seems like forgoing easy money to me.
 
If I was Sony, I would release it at the same price point as the current PS4 and just stop making the PS4.

I imagine AMD have made Sony a very attractive offer on the hardware to even make this a possibility so why not. No one is going to buy the less powerful PS4 anyway if they have a choice.

I don't think they'll go as far as stopping PS4 production right away. But I do think they will release this new version as a slightly smaller more powerful console for $100 more and not do any redesign for original PS4 and it will make the decision easy for consumers.
 
Status
Not open for further replies.
Back
Top