I think what's possibly more important is how low of a price point can they get the PS4 down to. Maybe we'll never see $99 again but maybe we will? I can see Sony wanting to ride out a year or two of high volume mainstream sales. They may not be the most profitable sales, but the mindshare into the average home is probably worth it.
Predict: Next gen console tech (9th iteration and 10th iteration edition) [2014 - 2017]
- Thread starter Shifty Geezer
- Start date
- Status
I think what's possibly more important is how low of a price point can they get the PS4 down to. Maybe we'll never see $99 again but maybe we will? I can see Sony wanting to ride out a year or two of high volume mainstream sales. They may not be the most profitable sales, but the mindshare into the average home is probably worth it.
There's more to these consoles than ever (Wifi, BT, HDD) while inflation has devalued what $99 can get you. Only was $99 is being reached is a loss leader. It's questionable how much people wanting a $99 console would be willing to spend on software to make up that loss.
Prophecy2k
Veteran
There's also the profit factor to bear in mind. Sony is doing so well with the PS4 at the current entry price, that there's little justification for them dropping the price down to $99 (even if they could without taking a loss on each unit). Why sell at $99 at cost when you can sell at $150 or $175 and make a healthy profit on each box. Current Sony is all about making money, given the success of their Playstation division over their other arms.
Prophecy2k
Veteran
I suspect Sony will be planning for a late 2019 launch, partly because they envision MS gunning for a 2020 launch. A PS5 on the market for a year before MS can get a Scorpio follow-up out the door will really cement for Sony a global lead a second time.
Some of that may depend on how Scorpio performs / ends up being perceived.
If it can show that coming out a year later in roughly the same ball park price (if that's the case) that it looks and performs a lot better. Then people may project that situation onto ps5 and scorpio 2.
If games don't look or perform much better on scorpio, then people may just jump on ps5 as why bother wait.
I think that what really makes the difference is the 4k assets for scorpio compared to the rest of the consoles.
PS5 and year later Scorpio 2, I'm not sure what will be the big differentiator between the two, yea power, but will that be enough to show a as big a difference as asset quality?
Scorpio 2 could come out with double the memory, but if the assets are 4k on both console, will it make much difference, doubt it.
If that's the case maybe MS will launch Scorpio 2 same year as ps5. Sony launched 4pro year earlier, maybe ms will launch Scorpio 2 year early this time around.
For MS if their not starting a next gen, then it's not really cutting the life of Scorpio, they'll just be moving it to budget device bit earlier.
Silencerx98
Newcomer
Sony can push the PS5 out the same year as the Scorpio successor, or hell, a year later and the PS5 will still have a global lead, LOL. Just look at the PS3, for instance. A year late to the market but eventually managed to catch up and surpass the 360 in sales, even though the PS3 was received negatively prior to launch. In addition, Microsoft was trying to expand the brand to a worldwide appeal in most of the 360 lifespan, can't say the same today. If Microsoft continues ignoring most of the countries besides US and UK, and they have consistently shown this gen they intend to, an XBox console will never be the market leader. The only exception is if Sony somehow fucks up even worse than with the PS3 and ditto for Nintendo in regards to the Wii U.
I know. But then again I don't see a reason for Sony to be in a position to be forced to pull the trigger by 2019. It's doubtful that MS can make up that level of lost ground over the past 4 years. No new Nintendo hardware on the horizon. Not really a lot of pressure riding on Sony imo. Might as well just work on a really good piece of hardware for PS5.
Prophecy2k
Veteran
Are we safe to assume this, especially after PS4 Pro and Scorpio?
Wouldn't there be an incentive for MS, for example, to try to go with a more aggressive design to 1up Sony on the performance marketing bullet point?
It might not help them in terms of global sales, but it might be enough for them to win back their main territories, UK and US, provided performance is the sticking point for consumers in those regions.
I think we are at a point where there are obvious launch times for new technology given constraints on power and price. Waiting a year might give you a decent jump in performance in the past but now, it may not, so both parties will look to launch when they can maximize the use of existing technology and minimize missed potential of new technology. Unless either party is willing to pay a lot more for custom things or they move from using AMD tech, I don't see them getting substantially different systems aside from maybe some adjustments integrated into a primarily AMD designed SoC. AMD itself is bound by their own R&D schedule which is bound by the founderies' processes.
It all comes down to if waiting a year gets you 10% more, neither will wait but if something big will come from waiting 1 year, both will look to wait. I think the ps4 pro and scorpio are going to be small bumps in the grand scheme of things regarding launch schedule. Of course tech will still need to make sacrifices to market conditions but to me, it seems like tech is slowed down to the point where new consoles will simply have to wait for a new wave of tech and new tech are grouped such that there is clear entry points to maximize performance while minimizing power and still delivering target performance at the specific power.
Is it a sticking point though?
We can probably assume that the customers who were most preoccupied with performance went with the PS4 this generation. However, even in that selection of consumers, the interest in the PS4 Pro has been Quite Modest. So it is difficult to see that Scorpio will cause any major shift - where would it come from? The most performance oriented consumers got a PS4, and may belong to the small subgroup among PS4 owners that upgraded to PS4Pro. Will they jump ship to Scorpio for another small hike in IQ? Or will the XBOne stalwarts that weren't interested enough in performance to be swayed by either the PS4 or the PS4Pro suddenly migrate en masse to Scorpio? Neither scenario seems likely.
In other news, in my country I yesterday saw the XBoxOne Slim on sale for just under $150. Don't know if they sold any. But they are trying.
Megadrive1988
Veteran
I'm thinking that because the first Zen / RyZen taped out around late 2015 (I could be wrong) that PS5 and a true next-gen Xbox will both be based around Zen 2 architecture, a low power version.
8 cores, 16 threads, somewhere around 3+ GHz.
Navi GPU architecture with some features from whatever follows Navi. Sony will want a 10x leap in performance above the original PS4, so around 18 TFLOPs.
RAM might (hopefully) be HBM3 because it will cost less to produce than HBM2, although not cheap. The low-cost HBM that's being made for consumer electronics probably won't have enough bandwidth.
So either HBM2 or HBM3, and I don't see Sony going for less than 32GB of whatever flavor of HBM they choose.
Although some expect Fall 2019 for PS5 release, I think Sony will decide to stretch things out 1 more year, for a number of reasons. Developers still working on PS4 games in late 2018/2019. Especially parts of Final Fantasy VII Remake. The process for 7nm will matureg, get better and faster. Many other factors, not just hardware but also more importantly, software. So I think Sony will wait until late 2019 or early 2020 to *announce* PS5. Blowout at E3 2020 - Launch September 2020 in NA and EU for the 25th anniversary of PlayStation's release. (I know the 25th anniversary in Japan will be December 3rd, 2019, so between then and Feb. 2020 would be a good time to announce PS5 at a PlayStation Meeting.
Ultimately, this is in the hands of Mark Cerny and the PlayStation architecture team -- So nevermind my predictions, he'll call the shots on what PS5 is.
8 cores, 16 threads, somewhere around 3+ GHz.
Navi GPU architecture with some features from whatever follows Navi. Sony will want a 10x leap in performance above the original PS4, so around 18 TFLOPs.
RAM might (hopefully) be HBM3 because it will cost less to produce than HBM2, although not cheap. The low-cost HBM that's being made for consumer electronics probably won't have enough bandwidth.
So either HBM2 or HBM3, and I don't see Sony going for less than 32GB of whatever flavor of HBM they choose.
Although some expect Fall 2019 for PS5 release, I think Sony will decide to stretch things out 1 more year, for a number of reasons. Developers still working on PS4 games in late 2018/2019. Especially parts of Final Fantasy VII Remake. The process for 7nm will matureg, get better and faster. Many other factors, not just hardware but also more importantly, software. So I think Sony will wait until late 2019 or early 2020 to *announce* PS5. Blowout at E3 2020 - Launch September 2020 in NA and EU for the 25th anniversary of PlayStation's release. (I know the 25th anniversary in Japan will be December 3rd, 2019, so between then and Feb. 2020 would be a good time to announce PS5 at a PlayStation Meeting.
Ultimately, this is in the hands of Mark Cerny and the PlayStation architecture team -- So nevermind my predictions, he'll call the shots on what PS5 is.
Prophecy2k
Veteran
I would agree somewhat, but I would caution on these assumptions. The same things were essentially said before this gen, and yet we still got two consoles with significant differences in design and performance (i.e. PS4 and XB1), despite all the same circumstances appearing true.
My question earlier also referred more towards the "given constraints on power and price" part, and not that either MS or Sony would wait a year or not for better technology to become available. Even with the same technology options available and with both vendors designing their consoles with the same hw vendor (AMD), if say Sony designed a box for a given set of constraints on power and price, say they targeted a $399 launch price, and MS designed for a $499 price, I could very well see them both coming up with designs that were significantly different again in features and performance, despite launching in the same period.
Whilst I agree with the premise somewhat, I don't think we can use PS4 Pro uptake as a barometer for anything. After all a mid-gen console refresh in this manner is entirely unprecedented, and more importantly, there's a world of difference between a mid-gen refresh of the same hw and a next-generation console.
I do agree that there isn't really enough data to determine either way whether power is a real sticking point for early adopters of console hw at the start of a new generation. At the same time, however, this is irrelevant. The point isn't whether power is or isn't indeed a sticking point, rather is MS or Sony convinced that it is? As my suspicion, given the design of Scorpio and the way MS were willing to release a year later to get that level of performance, at least starts to suggest that MS think that power is important to the hardcore console gamer... and assuming they do, will this factor into their design goals for their next-box?
You definitely have a strong point in that the central issue is whether or not having "the most powerful" console is considered critical by MS and Sony, and whether they feel that it is more important than, say, launching a year earlier and build early adopter volume, or hitting a lower price point for entry.
However, I'd say that the waters are significantly muddled as far as generations go. I think we can safely assume that Sony for example already has firm plans for the PS5. Given the PS4 and PS4.5 why would they design an incompatible PS5? Rather, the generational boundary is likely to be that they no longer make PS4 support mandatory for new titles, but won't business sense pretty much dictate that early titles be backwards compatible as well? This, on the other hand, will tend to make the PS4 to PS5 line pretty blurry and the start of a new generation less compelling for early adopters. (This is even more the case for a potential Scorpio successor.)
It's a bit of a new ball game.
HBRU
Regular
The main thing to be understood is that Sony (but also MS) earn on the software and marginally on the hardware... both have therefore the maximum interest to distribute the greatest number of copies of games for each console (new or old)... and therefore IMHO the PS4 (also seen the level reached by the marvelous Horizon Zero) will still have very very long life. Moreover the productive processes mature (if they mature) now very very slowly... and still it is not even known if there will be a viable productive process beyond the 7 nanometer, and the 7 same nanometer are at risk and all could also be stopped at 10. However in 2019 holyday a new Sony console will arrive IMHO... but that PS5 will totally ps4 backward compatible (and therefore it will also be born somehow old... lets face it). Produced at 10 nanometer, it will IMHO be a - further doubling of ps4-pro and therefore 72 CUs increased 20% in frequency and with the advanced functionalities of Vega, specially -the most interesting- the unified management of Cache / RAM / HD. So GPU 10 Teraflopa, CPU 16 cores improved Jaguar (not more in cluster from 4... but jaguars “made” single) @ 2,7 ghz, SSD 512 Gb, 16 Gb RAM GGDR6@512 GB / s). They will probably do a version of it with a reader blue ray 4k and one without any reader. Unfortunately for MS this PS5 will sweep out Scorpio even if initially it will cost 500 $ for the version with reader and 400 $ without. It will consume max 200 watts... over the 200 watts I don't think that is a good thing for a console.
One thing I fear about next gen consoles is that not being able to hit the same kind of traditional graphical leap every new gen provides, if we assume 4k/30fps is the baseline. It basically means you need a 4x increase in resolution while drastically improve the quality of assets, lighting, shadows, shaders, physics or whatever at the same time. I think we're looking at a 9-10x increase in power at the minimum and I don't see either MS or Sony hitting a 17tf + console with a reasonable price. Can 7nm save us?
Short answer:No
Longer answer - it depends.
At the same power draw, TSMC 7nm marketing claims that you can increase complexity almost three times vs. 16nm FinFet. That would actually be very good, if it turns out to be true, and for a GPU could translate almost directly into three times the performance IF other conditions are met. Additionally, Sony could push die size and power draw higher than the standards set by the current generation (approx. 160W for PS4 Pro)
Will they though?
Depends on their cost, form factor and living room noise targets.
Also, they need to take corresponding steps in terms of memory speed and size.
So it is not that your ad hoc target of 17+ TF is unreachable, the question is if it makes sense from a total market analysis viewpoint. Only Sony can say. But I'd reckon a 450mm2 7nm SoC, with a total device power draw of somewhere around 250W as unlikely, which is in the ballpark of what would be required. There are other ways to skin that cat that may be more palatable. Because I agree, IF Sony wants to introduce a clear generational step, they should preferably move a factor of four out from the PS4pro in terms of ALU capabilites and memory size and speed.
Their strategy for the future is not that clear though, maybe it makes more sense for the market leader to emphasize a low threshold for current owners to migrate forward.
Nope. With every node now, the power and performance scaling isn't as good. I doubt going from 14/16FF to 7nm will be a 2X increase in both. The differences have to be made up for in other areas. HBM being one as it gives better bandwidth at a much lower power thus allowing the saved power to be used for compute (and hopefully not on cost savings). It wouldn't surpise me to see more dedicated/specialized silicon return as well.
I'm only expecting a 10-12 TF next gen because of this.
Prophecy2k
Veteran
I know this is bordering a little on off topic, but will there be any potential die shrinks available for a next-gen console launching at the 7nm process node?
Will this likely impact the design, i.e. perhaps a motivation to future proof by going "bigger" or even going "smaller" depending on whether you think future-proofing performance is more important than future-proofing price?
Will this likely impact the design, i.e. perhaps a motivation to future proof by going "bigger" or even going "smaller" depending on whether you think future-proofing performance is more important than future-proofing price?
- Status