Xbox Series... M?

[Deleted member 11852]

It would be odd as the majority of games support variable resolutions. So even if the mobile device needed to render at a lower resolution it should be trivial to add in the lower bound resolution. But like I said it seems odd since AMD is so close to the series s in a mobile apu now. I think 5nm would let them hit it

It isn't as trivial to revisit and change older games as you might think. The reason more older games have not received 60fps bumps on the new consoles is because to republish updates you need to rebuild games using the latest console APIs and they often require the latest middleware, and that sometimes brings incompatibilities with older games which is more work.

It's less challenging on games already still in support and new games to come.

 

[Jay]

Jez seems to think that such a device may require dedicated support. Otherwise it will have to play games from the cloud. Granted he also says he knows nothing of such of device. He's just talking hypothetical. LOL

https://twitter.com/x/status/1542589533933473799

https://twitter.com/x/status/1542591102129143813

Tommy McClain

I agree with the people he spoke to at MS saying a portable XSS isn't viable. He mentioned that in a podcast.

His idea of subset of native games makes no sense.
A portable 1S with very good screen, battery life and controls would make lot more sense.
Play 90% of games from now going back to og xbox. So big catalogue to choose from.
XS games streamed onto a proper handheld.

 

[Silent_Buddha]

It isn't as trivial to revisit and change older games as you might think. The reason more older games have not received 60fps bumps on the new consoles is because to republish updates you need to rebuild games using the latest console APIs and they often require the latest middleware, and that sometimes brings incompatibilities with older games which is more work.

It's less challenging on games already still in support and new games to come.

Not that I think any developer would do this, but here's a scenario where it wouldn't be so arduous.

On PC, games are scalable via user configuration of options available for the game. You could do something similar on console while still keeping the simplicity of console gaming.

First, you'd have to have a console maker that commits to support for whatever APIs and environments are used across a range of consoles.

Then, ensure that your game is scalable as is usually the case on PC. Then, instead of having numerous gamer accessible configuration options (to preserve the "console" experience), have those options available to the console maker.

Console maker could then scale settings up (or down) whenever they release a new console. They can go so far as to enable extra features that couldn't make it into a console release (but exist, for example, in a PC release) due to console performance profile when a new more powerful console is released. They could also then reduce or even disable rendering features if they were to release a less capable console in the future. A practical downward reduction limit would be the PC versions lowest configuration settings.

For any developer that is going to release a PC version most of that work is already going to be done due to the existence of a PC version. Granted some extra work may need to be done on console to expose scalability settings and allow the game to scale on console. But once a game is release, if the developer will not or cannot update the game for new hardware for whatever reason, the console maker could then opt to do that if they wanted at minimal time and cost investment.

Obviously, this would be more difficult for console developers who currently do not do the PC version of a game in house and instead rely on 3rd party developers to port their game. IE - game developers who don't know how to scale their engine and rely on a 3rd party to do it for them.

Regards,
SB

 

[Deleted member 11852]

Not that I think any developer would do this, but here's a scenario where it wouldn't be so arduous.

I think you missed my point. It's not about planning for flexibility in the future, it's about retrospectively adding flexibility for a new performance profile (Series M) in games already released - the games that people may want to play.

Most of the last-gen games I own on Series X and PS5 still run at 30fps. A bunch do run at 60fps, either through game patches or Xbox's FPS Booster, but most - the vast, vast majority of older games - do not. Including games that would benefit like Red Dead Redemption 2 which Rockstar are still updating and supporting.

 

[eastmen]

It isn't as trivial to revisit and change older games as you might think. The reason more older games have not received 60fps bumps on the new consoles is because to republish updates you need to rebuild games using the latest console APIs and they often require the latest middleware, and that sometimes brings incompatibilities with older games which is more work.

It's less challenging on games already still in support and new games to come.

I think you missed my point. It's not about planning for flexibility in the future, it's about retrospectively adding flexibility for a new performance profile (Series M) in games already released - the games that people may want to play.

Most of the last-gen games I own on Series X and PS5 still run at 30fps. A bunch do run at 60fps, either through game patches or Xbox's FPS Booster, but most - the vast, vast majority of older games - do not. Including games that would benefit like Red Dead Redemption 2 which Rockstar are still updating and supporting.

What old games are we talking about. If we are talking about Xbox one games. Those games have to run on the original xbox one from 2013. We already see the steam deck what was supposed to be a 2021 SOC from amd already running games at better than ps4 levels. So just using the built in xbox bc those games should run fine on any zen2/rdna2 product since the work is already done for the series s and x.

I think the problem with targeting XBO is that the generation is basicly over. This is the last gasps of it and by the end of 2023 (in which i envision is the earliest we would get a series m) the cross gen games will be almost completely done. So now you have a mobile console just releasing that is going to require custom ports almost from the get go.

However van gogh on 7nm already supports
Zen 2 4c/8t 2.4-3.5ghz
RDNA 2 8 CU 1-1.6ghz

the 6nm 6800u is
Zen 3+ 8c/16t 2.7ghz-4ghz
RDNA 2 12 cu's at 2200mhz

xbox series s is
zen 2 8c/16t @ 3.4-3.6ghz
rdna 2 20C @ 1.56 ghz

So AMD is almost there. It would make little sense at almost half way through a new generation to put out a last gen machine. If we assume 2023/2024 for a series m then 5nm should be available. I am sure that amd would be able to get to either 20cus @1.56ghz or a smaller amount of cus at faster speeds equaling the same performance. in terms of cpu clock speeds they are pretty close to the 3.4- 3.6ghz of the zen 2. So again 5nm might enable that.

 

[Deleted member 11852]

What old games are we talking about. If we are talking about Xbox one games. Those games have to run on the original xbox one from 2013. We already see the steam deck what was supposed to be a 2021 SOC from amd already running games at better than ps4 levels. So just using the built in xbox bc those games should run fine on any zen2/rdna2 product since the work is already done for the series s and x.

Well.. all of them.

For example, as we're seeing a lot more this generation, reducing the resolution means cranking the framerate. Suppose Series M has a 720p screen, then a modern, higher-clocked chipset, could offer Xbox One-era games for a lower-reosluton screen at 60fps - but only where the game was designed to support that, which few games were. When One X appears, the technical stretch was not framerate but resolution. More resolution on a handheld device may not be the priority over framerate.

But perhaps I misunderstand you.. ?

 

[Jay]

I think the problem with targeting XBO is that the generation is basicly over. This is the last gasps of it and by the end of 2023 (in which i envision is the earliest we would get a series m) the cross gen games will be almost completely done. So now you have a mobile console just releasing that is going to require custom ports almost from the get go.

Taking this slightly out if context, for my own purposes.

Where as I don't see that as a problem as you wouldn't be targeting it. It would stream current gen only games.
Reason for not being streaming only is for times you don't have decent enough net access, and so have access to large library of games.
Apart from that or streaming only device I personally don't see it happening at all.

Would mainly be for people that don't like add ons like the backbone and don't think they are good enough.

Technically xbox is all in on streaming for mobile.

 

[eastmen]

Taking this slightly out if context, for my own purposes.

Where as I don't see that as a problem as you wouldn't be targeting it. It would stream current gen only games.
Reason for not being streaming only is for times you don't have decent enough net access, and so have access to large library of games.
Apart from that or streaming only device I personally don't see it happening at all.

Would mainly be for people that don't like add ons like the backbone and don't think they are good enough.

Technically xbox is all in on streaming for mobile.

I am assuming you mean it would play all previous gen games natively and current gen only streamed ? Like I said I think that is a huge mistake. By the time such a device came out the current gen would be halfway over if its a 7 year cycle like the past one (2013-2020)

Xcloud works on a phone because you pay for a dedicated LTE/5G connection. While in this day and age in America I can get wifi almost anywhere I go, it is normally slow wifi due to it having to serve multiple customers. I have tried to play xcloud at a starbucks and mc donalds before and even at hotels sometimes its playable and sometimes its horrible. Then its even worse for transit even with an 5G connection moving so quickly in a car or train is to much for the service most of the time and whenever you hit a tower handoff the streaming experience is bad

Xbox is only all in for streaming on mobile because they don't have any other products currently and its easy to just have xcloud on all devices since it takes just a front end website or app to really support it.

Like I said , AMD is very close to having an APU at series s performance levels at the proper power requirements for a mobile device. I think these half measures will only hamper whatever would come out.

Another thing we haven't really discussed is new content. So sure a lot of people would be happy getting access to the xbox one or ps4 library of games. But then what? IF all new releases of games are xbs/ps5 only your handheld is going to require custom games moving forward. If its a straight up xbo or ps4 in mobile form factor it would require keeping old performance targets alive going another 5 years +. Looking at the switch you can see its lack of high end third party games and you can see the lazy cloud only ports just start to role in. On the flip side if you say well the old content is playable locally but all new games from the current generation is stream only , then you have the problems i mentioned above. Maybe this is mitigated by selling the device with requiring at 5g connection from a provider but I don't see people rushing to do that.


At the end of the day you put an amd apu on 5nm with rdna 2 and whatever zen version you want. Dial in performance to xbox series s performance levels. Put in a 800p or 1080p screen and call it a day. If the surface team has a hand in designing the system you can get a really nice design that maximizes battery size. They should be able to put in a larger battery than the steam deck which has a 40whr battery. Looking at the internals of the steam deck there could have been more space optimization and even the simple fact of loosing the trackpads on such a device and would have opened up more space for such a thing.




The original Nintendo switch had a 2.5-6.5 hour battery life https://en-americas-support.nintend...46835/~/how-long-does-the-battery-charge-last So microsoft hitting that target should be fine also. I think they could do that. The steam deck could have with a larger battery.

 

[Johnny Awesome]

@eastmen You make the best argument yet for not simply relying on xCloud yet. It might be too early for such a thing. Long term I still think this is where things are headed, but maybe not so soon.

 

[milk]

The original Nintendo switch had a 2.5-6.5 hour battery life. So microsoft hitting that target should be fine also. I think they could do that. The steam deck could have with a larger battery.

I suppose as long as the thinf can support Series S tier games at the same or nearly the same performance out of the box, even at a low batery life, they can still allow devs to offer an optional "extended batery" mode for their games. Devs who can't be bothered don't need to, those who can, do. I think its a good compromise.

 

[eastmen]

@eastmen You make the best argument yet for not simply relying on xCloud yet. It might be too early for such a thing. Long term I still think this is where things are headed, but maybe not so soon.

I don't think we will ever get to the point of xcloud or cloud gaming being the main thing. It's to easy for MS or Sony to take whatever year it is zen and radeon apu and put it in a box and sell it. They will still need new designs for the cloud gaming anyway. So why not let the end user bear both the cost of the hardware and the cost of the service .

I think we also have to look at the success of things like the switch. Parents don't want to pay $15+ a month for 5g for kids to play games. What's more is what happens when you start to get issues with the stream? Little jimmy will be crying like crazy if he looses fortnite on the trip to disney cause you drive through a tunnel or are too far from a tower to get good signal.

Using xcloud a lot for software flights I have learned a lot of limitations. For instance I visted my parents and played halo with my nephew on my galaxy note 9 casted to my tv worked great. Went literally across a street which would fit 3 cars next to each other to see my friend who was visting her parents. We tried it there and the game was unplayable. It was what 20 or so yards away? Looking into the future I don't know when stuff like this will be fixed or when it will be fixed and then you have huge areas of the states with extremely poor cell signal. We go fishing in PA by the poconos and I can barely make a phone call let alone stream a game.

I suppose as long as the thinf can support Series S tier games at the same or nearly the same performance out of the box, even at a low batery life, they can still allow devs to offer an optional "extended batery" mode for their games. Devs who can't be bothered don't need to, those who can, do. I think its a good compromise.

I am not sure how you would do an extended battery mode unless the devs program for it. But like I said the original switch got as low as 2.5 hours and trust me it was rough playing zelda with that level of battery life.

A portable series s should be able to get much better battery life playing xbox one and one x games. Series s games would just give less battery life. I would imagine it be a similar time line to the switch. Release on 5nm at series s performance targets and low battery life and then a refresh on 3nm would give you much better battery performance.

Of course like I said there is additional hardware inside the steam deck that wouldn't be in a series m. So they would be able to put in a bigger battery to compensate for power draw. Who knows they could get closer to 4 hours with the right battery size.

 

[milk]

I am not sure how you would do an extended battery mode unless the devs program for it.

That's what I meant. Let and encourage devs create an aditional Mobile-centric performance profile, but keep it optional.

 

[Johnny Awesome]

@eastmen Well, that's where our opinions differ: I think the tech will eventually favor xCloud and thin clients over local handheld power, but not for a little while yet.

 

[eastmen]

That's what I meant. Let and encourage devs create an aditional Mobile-centric performance profile, but keep it optional.

How so tho ? You are already reducing the resolution quite a lot. A 720p or 800p res is pretty low. You could I guess go lower to 540p and use something like FSA. I think the reason why a series M would work is because you just put the series s release on the portable. I think its a tougher sell to 3rd parties if they have to add to much additional work. Also like I said the switch sold just fine with only 2.5 hours of battery life.

@eastmen Well, that's where our opinions differ: I think the tech will eventually favor xCloud and thin clients over local handheld power, but not for a little while yet.

Maybe but I doubt it. We are only at the start of thin clients and I think that as time marches on its going to be more and more expensive for the end user. At the end of the day you are still going to need a device to actually run the thin client and those wear down and break. They become outdated with wifi , ethernet and so on. So you still have hardware turn over locally and then I have a feeling that thin client providers like MS and Google are going to be more reluctant to build out more and more capacity and continuously upgrade prior capacity.

If we look at microsoft their data centers in the USA are very limited and are almost always in low cost states. But to have xcloud really work they are going to need at least a data center in every state if not multiple.

Then there is the actual internet infrastructure. Not sure how it is in other countries but here we have Cablevision/altice and its the shittiest company I have ever had the displeasure of getting service from. We have a gigabit business line and we are constantly having connection issues. At least when we are having issues I can play single player games on my pc / steam deck or xbox. If it was all streaming i'd be shit out of luck. We have slowly been building a small ultra hd bluray collection because of this. When we see an interesting movie for $5 bucks on the platform we buy it and then if we have an issue with the internet we can simply pull out a disc. Man I miss fios.

Also cell service isn't much better. We have T-mobile and they will throttle your connection from 5g to LTE to 3g. It seems like all the big companies will do that if you hit certain caps. So I am not sure at least here in the us that xcloud or thin clients will ever really take off.

That's before we get into the end of moores law and micron shrinks and how do you continue providing more computational resources. At work I'd just as rather go back to a desktop computer to get more power.

 

[Deleted member 11852]

@eastmen Well, that's where our opinions differ: I think the tech will eventually favor xCloud and thin clients over local handheld power, but not for a little while yet.

Re-centralization of media services is absolutely inevitable. Once internet infrastructure gets to the point when geographic localisation isn't a barrier, and service is fast, low-latency and reliable, why would you want a box taking up space in your house, burning what is increasingly expensive electricity?

We're a way off, but it's definitely on the horizon.

 

[eastmen]

Re-centralization of media services is absolutely inevitable. Once internet infrastructure gets to the point when geographic localisation isn't a barrier, and service is fast, low-latency and reliable, why would you want a box taking up space in your house, burning what is increasingly expensive electricity?

We're a way off, but it's definitely on the horizon.

For movies and music sure. Music even lossless is not a huge file size for streaming , storage and processing. Video is kinda the same the sizes are bigger and will likely grow again with 8k over the next decade or two. But video games keep increasing in scope and fidelity. We are at the start of a generation but will people still want to stream xbox series x games in 2028 when new consoles are out with I dunno zen 6 and rdna 4 or 5 ? Do you think people would be happy with xbox one quality streams now ?

If we look at the xbox one generation and apply it to xcloud. 2013 would have been the start offering xbox one quality visuals. People would have been happy. But then in 2017 the xbox one x came out. Now in order to give xbox one x quality streams MS would have to reinvest again. But then 2020 the series consoles came and it requires another large investment.

For most gamers its just easier to drop money on the box if they get enough use out of it. I don't think that wont ever be the case. For casual gamers perhaps but not for more core gamers.

 

[milk]

I think the reason why a series M would work is because you just put the series s release on the portable. I think its a tougher sell to 3rd parties if they have to add to much additional work. Also like I said the switch sold just fine with only 2.5 hours of battery life.

Yes, man. That's why I said from the beginning that it would only work if it was entirely optional for the dev, for the sole purpose of drawing less power and stretching the batery life further. It has to be able to run Series S games unmodified if the dev doesn't bother to create the "mobile friendly" mode.

Please, read others carefully before replying, otherwise we get stuck in cyclical arguments.

 

[Deleted member 11852]

For movies and music sure. Music even lossless is not a huge file size for streaming , storage and processing. Video is kinda the same the sizes are bigger and will likely grow again with 8k over the next decade or two. But video games keep increasing in scope and fidelity. We are at the start of a generation but will people still want to stream xbox series x games in 2028 when new consoles are out with I dunno zen 6 and rdna 4 or 5 ? Do you think people would be happy with xbox one quality streams now ?

I feel like you are not differentiating between now and the future. Johnny Awesome and I are talking about the future.

Right now there are three technical barriers preventing wider-spread remote-gaming: 1) server localisation, i.e. servers needs to be fairly close geographically to the user to manage the second problem. 2) latency, plenty of console games running on the box in your living room already have poor latency between the gamer doing something on the gamepad and the game visually reflecting that. Now add in, sending the packets to the remote server, the server processing the input, a new frame being created, encoded and returned and displayed. 3) bandwidth, which is what is keeping both resolutions and fidelity (bitrate) of both video and audio far below what people can do on local boxes. Increasing bandwidth can also increase latency.

For most gamers its just easier to drop money on the box if they get enough use out of it. I don't think that wont ever be the case. For casual gamers perhaps but not for more core gamers.

This is what the music and movie industry thought. Younger people are more transient than they ever have been, they want access to content but there is a body of evidence that shows fewer than ever care about owning physical products.

 

[Proelite]

Yes, man. That's why I said from the beginning that it would only work if it was entirely optional for the dev, for the sole purpose of drawing less power and stretching the batery life further. It has to be able to run Series S games unmodified if the dev doesn't bother to create the "mobile friendly" mode.

Please, read others carefully before replying, otherwise we get stuck in cyclical arguments.

Jez Corden from Windows Central talked to Xbox HW folks and they said Series S is impossible in the forseeable future to put into a portable form. An Xbox handheld would need better battery life than all the PC handhelds including Steam Deck, so they might not be able to use the same node for the same perf envelope.

 

[eastmen]

I feel like you are not differentiating between now and the future. Johnny Awesome and I are talking about the future.

Right now there are three technical barriers preventing wider-spread remote-gaming: 1) server localisation, i.e. servers needs to be fairly close geographically to the user to manage the second problem. 2) latency, plenty of console games running on the box in your living room already have poor latency between the gamer doing something on the gamepad and the game visually reflecting that. Now add in, sending the packets to the remote server, the server processing the input, a new frame being created, encoded and returned and displayed. 3) bandwidth, which is what is keeping both resolutions and fidelity (bitrate) of both video and audio far below what people can do on local boxes. Increasing bandwidth can also increase latency.



This is what the music and movie industry thought. Younger people are more transient than they ever have been, they want access to content but there is a body of evidence that shows fewer than ever care about owning physical products.

But what is the future ? 100 years from now? 3g is what 15 or so years old now and we had data caps on it and now we have 5g and we have data caps on it. Why would we assume in the future we dont have data caps on mobile tech ?

That's before the fact that we are hitting peak towers. It's harder and harder to build out new towers and even with micro towers they are very hard to install in residential areas. No one wants a cell tower near their home. So you are still going to get range issues and tower hand off issues with cell conections.

With your reply. 1) Yes this is always going to be an issue. Because the servers needed are gigantic and unless we never improve from xbox series x they will continue to stay the same size or need to grow to keep up with demand 2) Yes some games on your console have poor latency. It is still better than that poor latency then having another layer of latency on top of it due to it having to hop around the internet to get to you. 3) Yes this is another problem and it wont change.


Again bringing music and movies into this is not a good comparison. Streaming music is already recorded and heavily compressed. Movies again are already recorded and heavily compressed. None of these are interactive. Video games are the only one of these where the user is constantly reacting to the visual stimulus and the game is reacting to the player reacting. This isn't going to change. That is why streaming will never take over completely. Even Microsoft agrees with it. They will put out another xbox because they still need to design an new generation of xcloud and if they are building hardware for xbox its easier to just put it in a box and have millions buy it a year. Cause at the end of the day 10m users a year buying a $300-$500 box to play the games is $300-$500 less that you have to spend waiting for that subscription cost to pay for that investment.