Sony PS6, Microsoft neXt Series - 10th gen console speculation [2020]

[xz321zx]

Hopefully technologies like DLSS3 frame generation (FG) matures and further research is made on decoupling viewport manipulation from game render (async reprojection). This would allow games to always output consistent vsynced 120fps or what ever locked fps for consistent image persistence and would enable the possibility to use display technologies like black frame insertion (BFI) or other strobing mechanisms reliably without crosstalk on any game. Even if underlying render loop only runs for example 45fps (without FG), async reprojection would make the input feel like 120fps.

Lots of unanswered questions and research to be done but my main point is next generation is more about innovation on various aspects in the full software stack than raw hardware breakthroughs.

would enable the possibility to use display technologies like black frame insertion (BFI) or other strobing mechanisms reliably without crosstalk on any game.

That's deeply wrong, the reason you get crosstalk and unusable below 90hz is because the display is updated line by line and the backlight is not. Because they fail to deliver on that for last ten years at least.

Plus this is hardly decoupled in the traditional sense (that's texel shading).

 

[Seanspeed]

Wouldn't have thought it would matter much, if Intel are serious about being a fab for everyone.

More interesting to me is if Intel are in a position to shop around a console SoC in the next few years. Mostly this is due to XeSS and dedicated ML / ray tracing being interesting features vs knowing nothing about how RDNA will evolve.

Not really sure it's a worthwhile business venture for Intel, before you get into risks.

I was thinking sensitivity on AMD's side more than Intel's side. I'm sure Intel wouldn't care, though the scale required would be pretty demanding.

As for an Intel SoC, that could definitely be interesting, but they really need to get their GPU's in better shape before I'd even think of them being a realistic option. Alchemist was an ok start, but the performance per mm² is actually kinda dreadful, and that matters a whole lot for a console.

 

[Cyan]

Mark Cerny: When making consoles, we're not trying to build low-cost PCs

PlayStation legend Mark Cerny discusses PS5, the challenges of making consoles, and his 42-year games industry career

www.gamesindustry.biz

Mark Cerny Deep Dives on Ray Tracing, Frame Rate & Hints at PS6

 

[function]

Mark Cerny: When making consoles, we're not trying to build low-cost PCs

PlayStation legend Mark Cerny discusses PS5, the challenges of making consoles, and his 42-year games industry career

www.gamesindustry.biz

I read the article a few weeks back. Mark Cerny continues to be a class act and very grounded, no bullshitting about being all foreseeing and all calculating. He's trying to balance out of the gate cost and performance against complexity and features that will take years and years to see use (arguably the case with some Xbox features).

"I have been very surprised by the degree to which developers are using ray-tracing," he begins. "Putting that in, that was a big decision and actually a rather late one. I thought that this is not going to get much use initially, but if we look at generations, and a generation is seven years or so, software is created for ten years, and so later in the lifecycle we will start seeing people using that technology. But instead we had launch titles that were taking advantage of it. I guess, having worked on games for consoles that were a bit difficult to get into, like the PlayStation 3, I can be a little skittish about very deep technology like that. But in this case, my guess as to how things would go was totally wrong. And I am so happy to have seen the early adoption of the technology."

This fits perfectly with Digital Foundry (I think it was @Dictator ) saying that PS5 did not initially have hardware RT, and that the hardware was delayed. It also fits with how PS5 doesn't appear to have any other RDNA2 features.

Expect PS5 Pro and PS6 to be based on contemporaneous GPUs (in the PS5's case Navi 10) but maybe with a couple of key features brought in early from AMD's GPU roadmap.

Also:

He points to an amusing video by Linus Tech Tips, which attempted to 'kill' the PlayStation 5 by building a $500 gaming PC that outperformed the console.

"They had to get a used motherboard," he says. "That was the only way that they could build a PlayStation 5 equivalent for a PlayStation 5 price. And if you're using used parts… well you can get a used PlayStation 5 for eBay for $300-something."

Good man. I love the PC and it's where I game. But so many "PC vs Console" comparisons involve re-using parts, buying a shitty kb&m, not getting a pad, and pirating Windows (or using some dodgy ass 3rd party Martian keys website).

 

[Cyan]

I read the article a few weeks back. Mark Cerny continues to be a class act and very grounded, no bullshitting about being all foreseeing and all calculating. He's trying to balance out of the gate cost and performance against complexity and features that will take years and years to see use (arguably the case with some Xbox features).

"I have been very surprised by the degree to which developers are using ray-tracing," he begins. "Putting that in, that was a big decision and actually a rather late one. I thought that this is not going to get much use initially, but if we look at generations, and a generation is seven years or so, software is created for ten years, and so later in the lifecycle we will start seeing people using that technology. But instead we had launch titles that were taking advantage of it. I guess, having worked on games for consoles that were a bit difficult to get into, like the PlayStation 3, I can be a little skittish about very deep technology like that. But in this case, my guess as to how things would go was totally wrong. And I am so happy to have seen the early adoption of the technology."

This fits perfectly with Digital Foundry (I think it was @Dictator ) saying that PS5 did not initially have hardware RT, and that the hardware was delayed. It also fits with how PS5 doesn't appear to have any other RDNA2 features.

Expect PS5 Pro and PS6 to be based on contemporaneous GPUs (in the PS5's case Navi 10) but maybe with a couple of key features brought in early from AMD's GPU roadmap.

Also:

He points to an amusing video by Linus Tech Tips, which attempted to 'kill' the PlayStation 5 by building a $500 gaming PC that outperformed the console.

"They had to get a used motherboard," he says. "That was the only way that they could build a PlayStation 5 equivalent for a PlayStation 5 price. And if you're using used parts… well you can get a used PlayStation 5 for eBay for $300-something."

Good man. I love the PC and it's where I game. But so many "PC vs Console" comparisons involve re-using parts, buying a shitty kb&m, not getting a pad, and pirating Windows (or using some dodgy ass 3rd party Martian keys website).

yup, nice observation. To match the PS5 in price and performance...., a brand new computer or Mini PC with decent components + Lossless Scaling Frame Generation x 3 plus Performance toggle enabled, would be more like it.

More fair without clickbait shenanigans.

 

[dobwal]

Wouldn't have thought it would matter much, if Intel are serious about being a fab for everyone.

More interesting to me is if Intel are in a position to shop around a console SoC in the next few years. Mostly this is due to XeSS and dedicated ML / ray tracing being interesting features vs knowing nothing about how RDNA will evolve.

Not really sure it's a worthwhile business venture for Intel, before you get into risks.

Intel's involvement in producing console System-on-Chips (SOCs) would extend beyond the direct profit margins from these components alone. Collaborating with major partners like Microsoft and Sony could benefit Intel in several strategic ways:

1. Economies of Scale: Mass production of SOCs for consoles would help Intel achieve economies of scale, reducing overall production costs. This cost efficiency can extend to other products using similar architectures.
2. Research and Development: Joint R&D efforts with partners like Microsoft and Sony can drive innovation and technological advancements. Shared R&D costs can also mitigate the financial burden of developing new technologies.
3. Market Influence: Being the provider for leading gaming consoles might enhance Intel's brand visibility and reputation in the GPU market. This can positively impact sales and adoption of Intel's other products, such as PC GPUs.
4. Technology Transfer: Technologies developed or optimized for consoles can potentially be adapted or transferred to other Intel products, further enhancing their competitiveness in the broader market.

 

[rntongo]

Mark Cerny: When making consoles, we're not trying to build low-cost PCs

PlayStation legend Mark Cerny discusses PS5, the challenges of making consoles, and his 42-year games industry career

www.gamesindustry.biz

Mark Cerny Deep Dives on Ray Tracing, Frame Rate & Hints at PS6

Great interview and thanks for sharing. Cerny makes some good points and he's authoritative on the topic. I think the PS5 pro is going to perform quite well because of this as well the PS6 will be a very compelling device. Sony is focused on the end goal and build the hw back from that. MS needs to pick a leaf and go back to this strategy which is how Xbox ended up being created anyway. Fully software led decisions and building back from that.

 

[trinibwoy]

I wonder what led to the inclusion of RT in the PS5 given Cerny’s skepticism about adoption. Was it a marketing checkbox or did it ultimately come from him?

On the other hand he mentions that consoles have an inherent advantage in introducing new tech as they’re not bound by existing standards. That’s true (relative to PC) so maybe he was for RT but was expecting it to bear fruit much later in the cycle.

 

[Liandry]

We're ot trying to build low-cost PCs, but in the end they are. LOL.

 

[iroboto]

We're ot trying to build low-cost PCs, but in the end they are. LOL.

there was never the possibility. They have so many factors working in harmony to try to bring the price down as low as possible by the platform holders. There isn't any method on the open market to be able to create a brand new PC that can cost the same as a console. Nor should it. It's never been an equivalent debate - the PC can do significantly more than a console can do, and the market prices for those components reflect that.

 

[Shifty Geezer]

Nor should it.

Indeed. This would only be possible if the consoles have a huge markup. Assembling a PC from parts means 1) higher costs per part as can't negotiate the deals for lots of parts console companies can and 2) profit margins for each step of the supply chain including retail middleman. It's pretty much a law-of-thermodynamics level impossibility.

 

[Lurkmass]

I wonder what led to the inclusion of RT in the PS5 given Cerny’s skepticism about adoption. Was it a marketing checkbox or did it ultimately come from him?

On the other hand he mentions that consoles have an inherent advantage in introducing new tech as they’re not bound by existing standards. That’s true (relative to PC) so maybe he was for RT but was expecting it to bear fruit much later in the cycle.

Their initial ideology/philosophy of "believing in generations" hasn't quite exactly panned out for the most part ...

RT integration was only able to spread as far it currently did in games because of the refusal of many publishers to either move on from previous platforms or make significant technical enhancements ...

 

[Liandry]

As I said before if console holders can't make console on same level as top PC or at least near top PC for $500 they should make two SKUs, one for $500 and one top for $700-800. And there is a lot of people who would've bought that consoles, me also. Or 3 SKUs. Low-end, middle range and high-end. Series S and Series X are good examples, but there is room to expand that strategy. Smartphones made that step a long ago, and people buy phines even for $1500.

 

[Jubei]

I agree there is no way you can match a PS5 in terms of hardware cost. But its also an unfair comparison. I dont pay 80 euros for a game on PC and i dont pay a monthly fee to play online

PSN sales are terrible in comparison to PC sales. I mostly buy physical because its the only way you get a decent price.

 

[iroboto]

As I said before if console holders can't make console on same level as top PC or at least near top PC for $500 they should make two SKUs, one for $500 and one top for $700-800. And there is a lot of people who would've bought that consoles, me also. Or 3 SKUs. Low-end, middle range and high-end. Series S and Series X are good examples, but there is room to expand that strategy. Smartphones made that step a long ago, and people buy phines even for $1500.

Comparatively, mobile phones have massive use cases compared to consoles, they only just do about everything now a days in the palm of your hand. There is no world leader that isn't conducting some form of work on these devices!

It's very hard to justify such a high end piece of hardware to only play games. In the PC space, super heavy GPUs are also used for a large array of professional tasks as well.

We can also see a trend of games no longer requiring the best graphics anymore, the younger generation of gamers are not driven by graphics as our generation is.

The economics of console hardware has always been about best bang for buck. Deficiencies in the hardware are expected to be made up through software optimization specifically tailored to extract the maximum amount of performance out of the configuration.

 

[Liandry]

Comparatively, mobile phones have massive use cases compared to consoles, they only just do about everything now a days in the palm of your hand. There is no world leader that isn't conducting some form of work on these devices!

Yes, but almost all use those for TicTok, Instagram, Facebook, WhatsApp and Telegram. Even top tier phones.

It's very hard to justify such a high end piece of hardware to only play games. In the PC space, super heavy GPUs are also used for a large array of professional tasks as well.

At least people who I know buy high-eng GPUs for games.

We can also see a trend of games no longer requiring the best graphics anymore, the younger generation of gamers are not driven by graphics as our generation is.

Our generation still is very large.

The economics of console hardware has always been about best bang for buck. Deficiencies in the hardware are expected to be made up through software optimization specifically tailored to extract the maximum amount of performance out of the configuration.

True, but I also said not $1500 as phones but $700-800. And also there could be cheaper SKUs. People should understand what all become more expensive and some company should make first step. There will be choice, buy cheap or buy expensive. Not buy console or buy PC, (and many people doesn't want buy PC), or wait half+ generation till Pro console is out. But that is my opinion.

 

[iroboto]

True, but I also said not $1500 as phones but $700-800. And also there could be cheaper SKUs. People should understand what all become more expensive and some company should make first step. There will be choice, buy cheap or buy expensive. Not buy console or buy PC, (and many people doesn't want buy PC), or wait half+ generation till Pro console is out. But that is my opinion.

Yea i agree, there's going to be a couple years of nuance still remaining in the console space like you suggest.

I think the general assumption (from the way the ABK merger turned out) is that when the technology and economics works out, people will just move to streaming video games. One way to lower costs is to share hardware. Now you won't be able to play on an airplane flight, but connectivity and bandwidth and latency continue to make large strides. Game streaming is currently in it's infancy, but over time it should follow the path of media streaming which, now matured, works quite well everywhere!

I mean, many of us here, leveraging Geforce Now; to me this is the future given the price points of hardware today. 1x 4080 gpu will buy me several years of the Geforce Now service. That's not including all the other hardware you'd have to buy as well, and potential upgrading of the hardware in the service.

I think where we are headed with consoles will be more and more accelerators; but once you hit a hardwall of energy consumption, cloud technologies become more attractive.

 

[Charlietus]

Yea i agree, there's going to be a couple years of nuance still remaining in the console space like you suggest.

I think the general assumption (from the way the ABK merger turned out) is that when the technology and economics works out, people will just move to streaming video games. One way to lower costs is to share hardware. Now you won't be able to play on an airplane flight, but connectivity and bandwidth and latency continue to make large strides. Game streaming is currently in it's infancy, but over time it should follow the path of media streaming which, now matured, works quite well everywhere!

I mean, many of us here, leveraging Geforce Now; to me this is the future given the price points of hardware today. 1x 4080 gpu will buy me several years of the Geforce Now service. That's not including all the other hardware you'd have to buy as well, and potential upgrading of the hardware in the service.

I think where we are headed with consoles will be more and more accelerators; but once you hit a hardwall of energy consumption, cloud technologies become more attractive.

Are latency reductions making great strides? Fibre optics connections are already almost running at the speed of light, and it's not like there are innovations on the horizon that will make that much better, and the 4080 tier already runs at 120 fps with reflex for the lowest possible latency. I tried the 4080 tier on my fiber connection with 2.5gb/s and 20-30 milliseconds of ping, and when I'm connected directly to the router it's alright, it's not bad. But the moment I try to use it on my laptop or my phone and I go in to another room of my house, it becomes unusable, which defeats the whole purpose for me. With a wifi 6E modern router.

It's not mass market ready, and it will not be for decades. It also feels like trying to solve a problem that doesn't exist, as in buying a console-pc-laptop.

 

[Mitchings]

We need quantum-entangled modems!

 

[iroboto]

Are latency reductions making great strides?

Yes. We are increasingly moving towards edge compute and more investment in this space will drive even lower latencies and more efficient use of bandwidth. It is a fundamental requirement to move the world towards real-time applications over cloud.

It is one of the core reasons why 5G networks latency is about 4-5ms. Research on 6G networks started 5 years ago and is expected it will be about 100us -1ms. Bandwidth up to 1Tb/s. these next generation mobile networks are completely designed around edge deployments to achieve the speeds that is required while keeping costs much more manageable for telcos.

As we move towards consistent 4K 8K video streaming, you can’t be hauling all that data from data centre to clients. The entire network will explode, there’s just not enough bandwidth to support it.

I can’t comment on your 6E issues. It’s a bit like trying to diagnose someone’s computer issues without seeing it. I have no idea how you setup your 6E and what devices are paired with it etc. any poorly implemented network will perform badly. The reason wired is still solid is because those headaches are removed entirely.

And consumer gear, to be frank, sucks. Major reliability problems.