Xbox Series S [XBSS] [Release November 10 2020]

[Deleted member 13524]

Well let me ask you a question. With the outlook of future node processes how low in price do you think all 3 consoles will drop over their life span ?

I tend to assume the larger chips are the ones that scale down in price the most, with smaller node processes. That would make it SeriesX > PS5 > SeriesS.
However, the SoC isn't the only component that needs to scale to make a cheaper console. The SeriesX has 25% more memory channels so its PCB is bound to be more complex than the PS5's, OTOH the PS5 has more expensive storage and prices for that can go either up or down throughout the next years.

But in the end, the PSU + assembly + distribution + etc. costs tend to be rather stagnant as the years go by and they're probably rather similar among all 3 consoles, so as all the SoCs drop in production cost the SeriesX and PS5 will have a larger headroom to go down in price.
That's not to say the SeriesX can ever go down to $200 like the SeriesS probably will, but once you have a $300 12TFLOPs + disk option vs. a $200 4TFLOPs diskless option, I don't know many situations where the $100 delta is worth saving.

Remember ps4 launched at $400 and sells for $300 right now 7 years after launch. Xbox one launched at $500 with kinect and now without an optical drive sells for $250 before the digital one was discontinued and the optical version $300. Ps4 pro launched in 2016 and is still $400 and xbox one x never officially dropped under $500 outside of one time sales bringing it to $400 I believe.

I think we should make a clear distinction between the first couple of years where Sony and Microsoft are trying to sell as many consoles as possible while subsidizing the console, and the later years where both companies are trying to make money out of their hardware.
The PS4 Slim didn't go below $300 because Sony figured out they wouldn't sell a lot more consoles if they lowered the price (though it did go significantly lower during spontaneous sales like black fridays and the like). Likewise to the OneX and PS4 Pro IMO.

You can read DF's interview with them where they talk about the improvements they added to XSS that will make it perform above its "4tflop" power thats inside of zen and navi vs jaguar and gcn.

That would be valid in the former gen, but a week from now we'll all be dealing with RDNA2 in the gaming department (eventually across all GPU ranges in the PC too), so I don't know how the "punching above 4TFLOPs" statement will hold.

 

[thicc_gaf]

The problem is that 5 nm is unlikely to get there for them. 5 nm Lockhart is likely to cost roughly the same as 7 nm Lockhart. Memory also isn't progressing in cost nor size nearly as fast as they would need to even think about Lockhart in a fire stick form factor. Hell, I'm doubtful they could even get 5 nm Lockhart + everything else it needs into an Apple TV sized box, although it's potentially possible if they go with some expensive cooling and an external power brick. Part of the reason the XBS-S is as large as it is, is due to the relatively simplistic yet large SOC cooling solution that they use. But that means it's also cheaper. For XBS-S, cost was far more important that trying to get it to be as small as they could make it.

The end result would be a smaller Lockhart based machine at 5 nm that would potentially cost a fair bit more than the XBS-S.

Regards,
SB

This is where I think a Surface-style offshoot would probably do best, since it'd already sell higher anyhow. Actually I don't see why MS would necessarily want to lock Series S or even Series X tech exclusively to the consoles; at some point I think the Surface devices could essentially spin off revisions of those chips and they could then combine that with affordability through All-Access if desired.

Make the marketing targeted to the Surface-tier markets, and they could just focus on scaling down costs for Series S as-is where applicable to try hitting the costs lower and lower. A Firestick-style Series S? I agree, that won't ever happen, but they don't need it to, either. Anything that's Xbox & USB stick-related is probably orientated towards Xcloud streaming, there's already rumors around this IIRC. That seems the way forward for MS to hit the extremely price-conscious mainstream markets (that and Xcloud app expansion integration with other devices).

Its all steps on the road to a fire stick like device. I don't think they expect it over night.

You can read DF's interview with them where they talk about the improvements they added to XSS that will make it perform above its "4tflop" power thats inside of zen and navi vs jaguar and gcn. Like i said improvements in navi 3/4 and zen 4/5 can allow a similar 4tflop console to perform better than the current xss. If they are targeting the same 4tflops or so of power but with a newer architecture they should be able to use less power and produce a smaller console.

Smaller APU that uses less power and heat to perform the same or better , infinity cache will allow them to use slower ram with a smaller bus and doing all this will allow for a smaller power supply and smaller cooling which leads a smaller sized console.

I definitely think some scaled-down form of IC (likely not 128 MB; they would still want to save some die space for other, new silicon I think, plus save on costs) could be utilized, but there is something the issue with costs on 5nm (or even 5nm EUVL) compared to 7nm. MS stressed quite a bit at Hot Chips how prices aren't scaling down with shift to smaller nodes, that trend is likely to continue.

They could continue to go light on the storage amount for example; 512 GB NAND will cost a fair deal less in 2023/2024 than it costs in 2020. They could probably stick with GDDR6, it should go down a bit in price over time. How much could that bring the BOM down though, is the question. The Series S is probably costing MS some $350 or so on production costs, right? Maybe even a bit more than that; I think some word was being had that they lose more on each Series S compared to each Series X, which might line up with the production allocation numbers.

There could be room for MS to do a fork with Series S at the refresh point; something like a console, and something like a portable tablet device. Or maybe something even a bit more than that. I have some ideas on this, since I've been thinking a lot about mid-gen and 10th-gen console specs but...I had to go back and rewrite a lot of stuff after getting a bit of feedback on the boards, mainly from function. Some of my other ideas were a bit....extremely bullish xD.

I'd worry more if they can sell them. Series S was available to preorder for months in some Eu countries like Germany and even UK where Xbox demand is a bit softer. Whereas XSX sold out quickly everywhere. And from what we know allocations went more heavily to XSX.

TBF, that's Germany and UK. MS lost a lot of marketshare across Europe this gen, they have to put in a lot of work to gain it back and then some. It's certainly possible, though.

Series S is really aimed more at the casual and mainstream types, they aren't really too big into preorders. If the systems are there in plentiful numbers on the holidays, a lot of people will seriously consider picking one up.

 

[eastmen]

I tend to assume the larger chips are the ones that scale down in price the most, with smaller node processes. That would make it SeriesX > PS5 > SeriesS.
However, the SoC isn't the only component that needs to scale to make a cheaper console. The SeriesX has 25% more memory channels so its PCB is bound to be more complex than the PS5's, OTOH the PS5 has more expensive storage and prices for that can go either up or down throughout the next years.

That's traditional scaling. However if you are targeting similar performance with a new apu design , ram and process that wont always be the case. Designing a 4tflop 5nm navi 4 / zen 5 will be much cheaper to product than the Series x @ 5nm . Just like how MS is able to put out a navi 2 + zen console with ray tracing support and an ssd for cheaper than they can make a one x.

But in the end, the PSU + assembly + distribution + etc. costs tend to be rather stagnant as the years go by and they're probably rather similar among all 3 consoles, so as all the SoCs drop in production cost the SeriesX and PS5 will have a larger headroom to go down in price.
That's not to say the SeriesX can ever go down to $200 like the SeriesS probably will, but once you have a $300 12TFLOPs + disk option vs. a $200 4TFLOPs diskless option, I don't know many situations where the $100 delta is worth saving.

Why would the ps4 @ 300 sell better than the ps4 pro @ 400 ?

5nm zen 4/5 navi 3/4 , cheaper ram , smaller bus running cooler with a smaller fan / heatsink / power supply and casing . The cheaper they get it the lower all access is. The easier it is to add it into promos. Buy a Surface studio ? Get a free XSS . Get Duo 2 ? Get a free XSS and so on and so forth. Further they can drive pricing down and the smaller the form factor the more likely people will buy it for multiple tvs. In my house we have 2 fire tv cubes , 2 fire tvs 4k and 4 fire tv sticks for all the tvs we have in the house . Those range from $40-100 and not every tv needs a fire tv cube.

Same here not every tv needs a XSX device. Remember MS wants to sell you games yes and game pass but they also want you to rent movies and stream netflix and other stuff through their devices. They want a family to rent a movie on an xbox or windows because it will play on all their tvs and computers and phones. Not buy it on another platform because they have them on their tvs and phones and consoles.

I think we should make a clear distinction between the first couple of years where Sony and Microsoft are trying to sell as many consoles as possible while subsidizing the console, and the later years where both companies are trying to make money out of their hardware.
The PS4 Slim didn't go below $300 because Sony figured out they wouldn't sell a lot more consoles if they lowered the price (though it did go significantly lower during spontaneous sales like black fridays and the like). Likewise to the OneX and PS4 Pro IMO.

I don't believe the ps4 slim or xbox one costs much less than the current selling prices and the one off sales were loss leaders designed to get big holiday sale numbers

That would be valid in the former gen, but a week from now we'll all be dealing with RDNA2 in the gaming department (eventually across all GPU ranges in the PC too), so I don't know how the "punching above 4TFLOPs" statement will hold.

They were specifically comparing to Xbox one X. I am saying will a future zen and navi or whatever is on the road map equally be able to punch above its weight class ? Will say a 2024 5nm newly designed zen and newly designed rdna (or something else) gpu targeting 4tflop out perform xbox series s ?

 

[BRiT]

Will say a 2024 5nm newly designed zen and newly designed rdna (or something else) gpu targeting 4tflop out perform xbox series s ?

Answer: Absolutely, if they can get an ML-based solution working across the board like DLSS 2023 revision.

 

[mpg1]

When you even look at the actual MSRP's of the last gen consoles it didn't drop a whole lot even after 7 years. You do get an extra 500GB and typically a gem bundled in but there wasn't a whole lot of savings overall.

I'd be surprised if there will be much of a budge on MSRP's at all this gen. Any savings will probably go to doubling storage a few years down the road....maybe start getting a game bundled in the next year or so. But I wouldn't be shocked if a lot of the MSRP's for each of the consoles models stay the same the entire gen.

 

[eastmen]

I definitely think some scaled-down form of IC (likely not 128 MB; they would still want to save some die space for other, new silicon I think, plus save on costs) could be utilized, but there is something the issue with costs on 5nm (or even 5nm EUVL) compared to 7nm. MS stressed quite a bit at Hot Chips how prices aren't scaling down with shift to smaller nodes, that trend is likely to continue.

They could continue to go light on the storage amount for example; 512 GB NAND will cost a fair deal less in 2023/2024 than it costs in 2020. They could probably stick with GDDR6, it should go down a bit in price over time. How much could that bring the BOM down though, is the question. The Series S is probably costing MS some $350 or so on production costs, right? Maybe even a bit more than that; I think some word was being had that they lose more on each Series S compared to each Series X, which might line up with the production allocation numbers.

There could be room for MS to do a fork with Series S at the refresh point; something like a console, and something like a portable tablet device. Or maybe something even a bit more than that. I have some ideas on this, since I've been thinking a lot about mid-gen and 10th-gen console specs but...I had to go back and rewrite a lot of stuff after getting a bit of feedback on the boards, mainly from function. Some of my other ideas were a bit....extremely bullish xD.


.

Yea i don't know how much IC they need to be effective at 1440p as the only IC we know of is a 128megs across the navi 2 line up so far. If 1440p could work on 64MB that would be a small amount of die space. By adding IC they can reduce ram speeds and the bus size further reducing costs. I believe the cost of series s is closer to $300 , my sources have said $200/$250 were price points that were on the table and I doubt they would have been if it was $350 to make


I think you guys are a bit limited in what you think of when you think of a console. Look at what MS has done in the pc space. They brought out the surface pro then surface book and laptop and studio and hub and they will continue to bring out new devices.

 

[thicc_gaf]

Yea i don't know how much IC they need to be effective at 1440p as the only IC we know of is a 128megs across the navi 2 line up so far. If 1440p could work on 64MB that would be a small amount of die space. By adding IC they can reduce ram speeds and the bus size further reducing costs. I believe the cost of series s is closer to $300 , my sources have said $200/$250 were price points that were on the table and I doubt they would have been if it was $350 to make


I think you guys are a bit limited in what you think of when you think of a console. Look at what MS has done in the pc space. They brought out the surface pro then surface book and laptop and studio and hub and they will continue to bring out new devices.

Well, if RDNA 2 GPUs are targeting 4K and have 128 MB IC, I assume a system targeting native 1440p would only need around 32 MB - 48 MB of it, keeping to scale. And still do all the things you are suggesting could be done.

Hearing that Series S was targeting $200/$250 is interesting because then I'm wondering why they went with $299? Is that to suggest Series S is being sold at-cost, or just about? If the reason they went with $299 is because of increase in RAM costs (that's the main reason they went with 16 GB of it in Series X instead of 20 GB), then does that also imply at some point they were contemplating $449 or even $399 for Series X?

Truthfully, it's not so much we're only choosing to look at consoles from a limited POV; I actually have some pretty neat ideas for Series S refresh hopefully I can post soon in the other speculation thread. But, we also have to keep in mind what the average gamer thinks of when it comes to a console, too. We're a niche, in that respect, relative to the wider mainstream market. What MS does with their Surface devices might work for that area but it might not work so well with the gaming space. Already between the Series S and X, while all the reports on "difficult development" for Series S are IMO overstated, it hasn't exactly been 100% smooth scaling, either. Even Phil Spencer's admitted that it's a bit of extra work.

Any additional SKU they throw into the product line that has a divergence in technical specs, means more systems for 3P developers (and 1P, tho they benefit from 1P-tier resources and funding direct from MS) to optimize against, which will cost time and monetary resources. The one thing that'll dictate how much MS pushes in this regard is how friendly Gamecore plays with offloading the grunt of platform-specific optimization for third parties. If it truly makes multi-spec'd SKU profiles relatively platform-agnostic in terms of optimizations, getting the typical work out of the way, I can see MS being more aggressive with different product variations in the Series line that more closely mimics what they do in the Surface area. If that happens to not be the case, expect them to be more conservative.

There's also to consider the fact that they would need to manufacture more of these Series-based models than the comparative Surface model variants, too, since consoles are sold at a loss and production in very high quantities helps to curb losses, particularly when yields are mature (when yields are bad at the beginning you're better on saving costs by just manufacturing less units).

 

[Silent_Buddha]

Yea i don't know how much IC they need to be effective at 1440p as the only IC we know of is a 128megs across the navi 2 line up so far. If 1440p could work on 64MB that would be a small amount of die space. By adding IC they can reduce ram speeds and the bus size further reducing costs. I believe the cost of series s is closer to $300 , my sources have said $200/$250 were price points that were on the table and I doubt they would have been if it was $350 to make


I think you guys are a bit limited in what you think of when you think of a console. Look at what MS has done in the pc space. They brought out the surface pro then surface book and laptop and studio and hub and they will continue to bring out new devices.

This point isn't that they can't do it, of they can. The point for devices in the XBS-S class or lower is cost. The Surface line are premium devices and command a premium price for the performance they offer. A "stick" or "TV" like device needs to be cheap (under 100 or 200 USD) and small (Apple TV or smaller).

With the cost structure of IC and new nodes, it's often cheaper to make a new design on a smaller node than it is to try to move your current IC down to a smaller node. This is why we aren't seeing those big cost reductions that we used to get with PS3/X360 gen and earlier gens.

Regards,
SB

 

[eastmen]

Well, if RDNA 2 GPUs are targeting 4K and have 128 MB IC, I assume a system targeting native 1440p would only need around 32 MB - 48 MB of it, keeping to scale. And still do all the things you are suggesting could be done.

I'm not sure what they would need. if your saying 48megs then perhaps thats all they need

Hearing that Series S was targeting $200/$250 is interesting because then I'm wondering why they went with $299? Is that to suggest Series S is being sold at-cost, or just about? If the reason they went with $299 is because of increase in RAM costs (that's the main reason they went with 16 GB of it in Series X instead of 20 GB), then does that also imply at some point they were contemplating $449 or even $399 for Series X?

My understanding as i've said before that MS was willing to go down to $200 on the xss if they needed too due to pricing competition. Sony went $400/$500 however and MS felt they didn't need too or perhaps decided to take less losses in the first year or so

Truthfully, it's not so much we're only choosing to look at consoles from a limited POV; I actually have some pretty neat ideas for Series S refresh hopefully I can post soon in the other speculation thread. But, we also have to keep in mind what the average gamer thinks of when it comes to a console, too. We're a niche, in that respect, relative to the wider mainstream market. What MS does with their Surface devices might work for that area but it might not work so well with the gaming space. Already between the Series S and X, while all the reports on "difficult development" for Series S are IMO overstated, it hasn't exactly been 100% smooth scaling, either. Even Phil Spencer's admitted that it's a bit of extra work.

anything is more work. MS took their time in rolling out the surface devices they launched with the rt then the pro . the rt failed and it took a few years for them to introduce another new category in their surface line

Any additional SKU they throw into the product line that has a divergence in technical specs, means more systems for 3P developers (and 1P, tho they benefit from 1P-tier resources and funding direct from MS) to optimize against, which will cost time and monetary resources. The one thing that'll dictate how much MS pushes in this regard is how friendly Gamecore plays with offloading the grunt of platform-specific optimization for third parties. If it truly makes multi-spec'd SKU profiles relatively platform-agnostic in terms of optimizations, getting the typical work out of the way, I can see MS being more aggressive with different product variations in the Series line that more closely mimics what they do in the Surface area. If that happens to not be the case, expect them to be more conservative.

It really depends on what the base is. If for example the Series S is the floor and anything future is more powerful at everything than the series S but not as powerful as the x they can simply take the dynamic resolution games and they will just hit their dynamic cap more often or all the time vs the S. The difficulty would be having to go under the S which i don't believe is in their plan

You've prob seen me bullish on a xss dell ufo type set up. I post about it all the time. 5nm zen 4/5 + navi 3 or 4 targeting 4tflops with infinity cache would be a great 1080p or 1440p hand hand held

Give the ability to to attach a surface key pad and run your microsoft 365 suite and you got a surface gaming device. surface xbox so to speak

There's also to consider the fact that they would need to manufacture more of these Series-based models than the comparative Surface model variants, too, since consoles are sold at a loss and production in very high quantities helps to curb losses, particularly when yields are mature (when yields are bad at the beginning you're better on saving costs by just manufacturing less units).

On the flip side surface devices are refreshed every 12-18 months or so. an xbox portable could last half a decade or more on the market. For instance once they get to stick size or fire cube size that product can continue to sell for 5 or 6 years with no additional R&D

This point isn't that they can't do it, of they can. The point for devices in the XBS-S class or lower is cost. The Surface line are premium devices and command a premium price for the performance they offer. A "stick" or "TV" like device needs to be cheap (under 100 or 200 USD) and small (Apple TV or smaller).

With the cost structure of IC and new nodes, it's often cheaper to make a new design on a smaller node than it is to try to move your current IC down to a smaller node. This is why we aren't seeing those big cost reductions that we used to get with PS3/X360 gen and earlier gens.

Regards,
SB

And that is why I believe they will reach fire tv / stick sizes with newer designs targeting XSS performance.

IF AMD is able to increase performance again 50% going from navi 2 to navi 3 and 50% going to navi 4 then it be easy to create an APU that is much smaller , uses less power and produces less heat. Add infinity cache and you could get away with slower ram and a smaller fsb. Your not going to get fire stick size right away but you will continue to get smaller and smaller consoles

https://www.techradar.com/news/amd-promises-rdna-3-gpus-will-be-as-big-a-leap-forward-as-big-navi

There is a reason Microsoft says its the smallest xbox to date

 

[thicc_gaf]

I'm not sure what they would need. if your saying 48megs then perhaps thats all they need

My understanding as i've said before that MS was willing to go down to $200 on the xss if they needed too due to pricing competition. Sony went $400/$500 however and MS felt they didn't need too or perhaps decided to take less losses in the first year or so

Ah okay I understand. So, that could be more indicative they'd be willing to bleed more on margins with it (MS can certainly afford to, much moreso than Sony ever could, especially now that MS proper seems to be supporting the Xbox division a lot more these days), but I don't know if it's indicative the Series S is too much cheaper on BOM production costs. I'd have to do some digging, but I think I recall a quote from someone at MS (prob Phil), they were saying that they are losing money on Series X and Series S, so I think it's fair to guess Series S on production costs is at least $310. It's probably a bit higher than that though, $350 isn't unreasonable.

anything is more work. MS took their time in rolling out the surface devices they launched with the rt then the pro . the rt failed and it took a few years for them to introduce another new category in their surface line

It really depends on what the base is. If for example the Series S is the floor and anything future is more powerful at everything than the series S but not as powerful as the x they can simply take the dynamic resolution games and they will just hit their dynamic cap more often or all the time vs the S. The difficulty would be having to go under the S which i don't believe is in their plan

I guess that could work, but I'm curious how the wider gaming market would respond to a range of Series systems if they're simply resolution scalers but otherwise still fit a stationary home console design profile. Guessing this type of approach would work best if you clearly dilienate the products in terms of use-cases, form-factors etc. Similar to differentiation of Surface devices, but I think in case of Series systems it'd have to be even more distinct.

You've prob seen me bullish on a xss dell ufo type set up. I post about it all the time. 5nm zen 4/5 + navi 3 or 4 targeting 4tflops with infinity cache would be a great 1080p or 1440p hand hand held

Give the ability to to attach a surface key pad and run your microsoft 365 suite and you got a surface gaming device. surface xbox so to speak

Yeah, that would be a good idea for a future Series device in the product ecosystem. It's what I've thought of as well for a portable-based Series system, but I think they could be even more bullish on it, and pursue a hybrid design in an even more innovative form factor modularity. I like the idea of running productivity software like 356 Suite on it, maybe that could be done through spinning the APU and design into a Surface device, as you say.

And that is why I believe they will reach fire tv / stick sizes with newer designs targeting XSS performance.

IF AMD is able to increase performance again 50% going from navi 2 to navi 3 and 50% going to navi 4 then it be easy to create an APU that is much smaller , uses less power and produces less heat. Add infinity cache and you could get away with slower ram and a smaller fsb. Your not going to get fire stick size right away but you will continue to get smaller and smaller consoles

https://www.techradar.com/news/amd-promises-rdna-3-gpus-will-be-as-big-a-leap-forward-as-big-navi

There is a reason Microsoft says its the smallest xbox to date

Dunno about AMD's claims for RDNA 3; personally I'm thinking they'll hit a snag at some point with pushing higher and higher GPU clocks but that is just me, and I don't think that snag will be hit until maybe RDNA 4 or 5, because no matter the architectural changes there's only so far you can keep pushing clocks while having SOME semblance for sane power consumption even when you hit say 3nm EUVL, and processes smaller than that are currently not even fleshed out or necessarily a thing at this time, or in VERY early planning and testing phases.

I do think you can eventually reach a system like Series S at, say, 25 or so watts of power consumption, sometime around 2024, but I don't know if it'd necessarily be "cheaper". But as you say, implementing some form of IC could let them skim down on main memory bandwidth and the such. Not for a Series S refresh particularly; games are still going to expect 10 GB to work with, and a certain type of bandwidth on the main memory side of things. Though maybe that won't be as big a problem as I picture it could be; MS redesigned the whole memory system for One X compared to the base XBO and XBO games still ran perfectly fine on it.

EDIT: Just remembered, that ease of translation is probably because MS partitioned games and apps into their own hypervisor layers, and the OS to yet its own hypervisor. Assuming the Series systems continue this, just much more refined.

 

[eastmen]

Ah okay I understand. So, that could be more indicative they'd be willing to bleed more on margins with it (MS can certainly afford to, much moreso than Sony ever could, especially now that MS proper seems to be supporting the Xbox division a lot more these days), but I don't know if it's indicative the Series S is too much cheaper on BOM production costs. I'd have to do some digging, but I think I recall a quote from someone at MS (prob Phil), they were saying that they are losing money on Series X and Series S, so I think it's fair to guess Series S on production costs is at least $310. It's probably a bit higher than that though, $350 isn't unreasonable.

you have to get the exact wording they used. Are they taking a loss on the series x and series as combined where the X is loosing money but the S is breaking even or maybe making a little but with the X cost they still end up loosing ? Also what are the projections of costs over the next x amount of months

I guess that could work, but I'm curious how the wider gaming market would respond to a range of Series systems if they're simply resolution scalers but otherwise still fit a stationary home console design profile. Guessing this type of approach would work best if you clearly dilienate the products in terms of use-cases, form-factors etc. Similar to differentiation of Surface devices, but I think in case of Series systems it'd have to be even more distinct.

Same way they took to the ps4 and ps4 pro or xbox one and xbox one x. Obviously the people who only have an xbox series s will be fine with it either playing in 1080p /1440p or will stream xcloud. I believe that MS will always have a $300 and $500 box (it might fluctuated a bit with price cuts maybe sometimes 200/400 before being replaced) but the hardware in them will be refreshed every few years to the newest architecture as that is the easiest way to improve performance and power profiles now.

That may mean in 2024 a Navi 4 and Zen 6 at that point on 5nm could produce a system on par or better (esp with ray tracing and ML) than the Xbox series X currently is but at a $300 price point and at a $500 price point maybe we are looking at a 20tflop+ machine . What would be the point of selling the current systems ?

At the same time as I've been saying you take that same navi 4 and zen 6 apu and you cut it down so it plays games at least as well as the current XSS. What would such an APU and console look like ? Well navi 4 would be the third try for amd at ray tracing. Performance there should be much better, if they are able to scale 50% or so performance improvements like they did from navi - navi 2 then perhaps you don't need to be as agressive with the clocks and memory speeds. That is the basis of what I believe they will do going forward , its what i've been saying for at least half a year and to me hot chips cemented it in my mind as what MS will do

Yeah, that would be a good idea for a future Series device in the product ecosystem. It's what I've thought of as well for a portable-based Series system, but I think they could be even more bullish on it, and pursue a hybrid design in an even more innovative form factor modularity. I like the idea of running productivity software like 356 Suite on it, maybe that could be done through spinning the APU and design into a Surface device, as you say.

I think they will be very careful in what it can do because they will want to have it part of but apart from the other surface devices. They will wnat you to get a surface pro and a surface xbox.

One thing i've talked about before is a usb c xbox attachment for future surfaces. As a customer you go out and you buy a surface go or pro or book or laptop and you already have the screen and battery and other functionality. So just put the xbox series s in a small portable enclosure and your good to go.

Dunno about AMD's claims for RDNA 3; personally I'm thinking they'll hit a snag at some point with pushing higher and higher GPU clocks but that is just me, and I don't think that snag will be hit until maybe RDNA 4 or 5, because no matter the architectural changes there's only so far you can keep pushing clocks while having SOME semblance for sane power consumption even when you hit say 3nm EUVL, and processes smaller than that are currently not even fleshed out or necessarily a thing at this time, or in VERY early planning and testing phases.

Maybe but thats a decade away or so i'd say. AMD will most likely get large jumps through 5nm at the very least. Remember RDNA is the first major architecture on the graphics side from them in a decade who knows what else they can improve in it

I do think you can eventually reach a system like Series S at, say, 25 or so watts of power consumption, sometime around 2024, but I don't know if it'd necessarily be "cheaper". But as you say, implementing some form of IC could let them skim down on main memory bandwidth and the such. Not for a Series S refresh particularly; games are still going to expect 10 GB to work with, and a certain type of bandwidth on the main memory side of things. Though maybe that won't be as big a problem as I picture it could be; MS redesigned the whole memory system for One X compared to the base XBO and XBO games still ran perfectly fine on it.

I think like a pc it wont really care. A pc doesn't care if you put a geforce 1080 or a 3080 in it, it will just run the code as well as it can. MS has the direct x layer which should make it easy for games to run fine. IC @ xbox series s clock speeds should provide enough bandwidth at 1440p i'd imagine. I'd also wonder if a 5nm navi 4 would be able to run at higher clocks while removing Rops and CUs and still perform as well as the rdna inside of the xbox series s.

Like i said i think the xbox series s is the base for Microsoft and what comes next will be faster than it while using newer tech. They go into detail of why simply dropping the micron process on a design wont work for price drops like it used too.

EDIT: Just remembered, that ease of translation is probably because MS partitioned games and apps into their own hypervisor layers, and the OS to yet its own hypervisor. Assuming the Series systems continue this, just much more refined.

yea most likely

 

[Xbat]

OTOH the PS5 has more expensive storage and prices for that can go either up or down throughout the next years.

Is it more expensive though? Serious question. There's less and isn't the speed because of the extra lanes ?

 

[liams]

Is it more expensive though? Serious question. There's less and isn't the speed because of the extra lanes ?

There is less storage overall but the PlayStation uses a higher number of discrete chips than xbox, I know its not completely analogous but you could think of the sony solution as a RAID 0 array over a bunch of smaller chips. The xbox uses a single storage chip. So it's a matter of the cost per gb per chip, the xbox only has to pay for the IC packaging once, the PlayStation multiple (irrc 8 times)

EDITED: got my raid levels mixed up

 

[Johnny Awesome]

My prediction is that the next Series S and Series X consoles come out in 2024 on mature 5nm @ $299 and $499 again, mainly concentrating on better RT and ML. The existing S will be $199 by then. They might not even bother with X @ $399.

A lot of this depends on market conditions though. The main reason for the One X, was because MS wanted to regain their position as "most powerful console". If the One had sold like hotcakes or as well as PS4, they likely wouldn't have bothered. The PS4 Pro was just a preemptive move by Sony when they learned about the One X.

 

[dobwal]

I can see this whole gen being manufactured at 7nm. There is no real cost saving going to a lower node anymore. Apple and the like plow through the newer nodes for greater transistor density and power savings, not cost savings.

If you are going to drop new hardware with better performance, you might as well release them as new-gen hardware. And have cross-gen titles exist over the span of a gen offering support to the previous gen.

 

[Johnny Awesome]

Well, we're essentially saying the same thing I think, unless you think they'll wait the whole 7 years for the next hardware. They might, depending on sales. If the XSS outsells the XSX, there's not as much incentive to keep pushing the high end.

 

[Rangers]

More comparisons are coming out where XSS is not even matching X1X. The latest is COD Black Ops Cold War where One X is 1800-2160P and Series S is 1080P.

I wonder what's the issue, besides maybe not optimizing. RAW Flops? The 128 bit bus (it's 128 right?). The 10GB RAM vs 12 in One X?

 

[BRiT]

In those games, are they locked resolutions and not dynamic to at least let it scale up?

 

[JPT]

I thought the XSS was an 1080p box, that was the whole point of MS making it. So why would it it match the One X is resolution when One X goes over 1080p?
I mean can the one x do 120fps @1080p with Dirt5?

 

[Arwin]

More comparisons are coming out where XSS is not even matching X1X. The latest is COD Black Ops Cold War where One X is 1800-2160P and Series S is 1080P.

I wonder what's the issue, besides maybe not optimizing. RAW Flops? The 128 bit bus (it's 128 right?). The 10GB RAM vs 12 in One X?

Assassins Creed: Valhalla is also 30fps where Series X is 60fps, that is also disappointing and a bit hard to understand, really.

Maybe it is just a lack of priority.