Business Approach Comparison Sony PS4 and Microsoft Xbox

[joker454]

The Xbox One project was hundreds of millions to potentially billions of dollars, so I would be curious how much of that is "mostly", especially since releasing a new console quickly will hamper the earlier console's ability to maintain volume and pricing.
Unless you expect it to immediately stop making Xbox Ones, which just means Microsoft is perpetually stuck with a console that never has a chance to recoup its up front cost.

It would be a continuation of the same platform, so all marketing related costs carry over. It would be fully backward compatible so software costs mostly carry over. It would be using new hardware funded by Amd because Amd has to be doing that anyways, so no need to spend another billion dollars, just go with Amd's latest and greatest. There's no need to commission new hardware in the typical sense, nothing new to build or create. Its another xb1 revision using similar but more current hardware. I'm not saying it would be free, but it wouldn't cost anywhere near as much as launching an all new and incompatible platform. They don't need to stop making xb1's anymore than Apple has to immediately stop making iPhone 4's after they release the iPhone 5. Xb1's can continue to linger around as the cheap model until demand wanes, then it can be eliminated. Kinect would carry over as is as well.

Bog standard silicon has pretty high up front costs. Those fearsome tablets use silicon chips that must sell hundreds of thousands or millions of units to make up for the design, validation, and manufacturing costs.

If you mean creating an even more powerful APU, for example, it is a new chip and will incur much of the cost that went into Durango.
One thing to note is that AMD's little back and forth with Sony, itself, and Microsoft did mean that some of hoped-for cost savings in R&D already happened.
I'm also not sure in this scenario if you are envisioning a rapid retirement of Durango when Durango Redux comes out.

Any new hardware means incurring new up front costs. Giving Durango something over a year to pay off its up-front costs doesn't seem to be consistent with Microsoft's publically stated plans, at any rate.

Many current tablets cost mucho dinero because they are still custom designs based around Arm, hence why they have more costs. So like Apple starts with Arm then spends a chunk of time and money to customize things to their needs. I don't think MS need to go with that approach for a hardware revision. In other words they don't have to sit down with Amd and discuss what to build, instead in 3 years just go with Amd's current design be it a gcn 3.0 or whatever apu. At the 6 year time frame maybe they need to go more hardcore and spend some money for a new design, but for a 3 year revision model I don't think they need to bother. Think of it as an "Xbox One S", a refresh to keep the platform fresh and interesting, and competitive with advances on tablets, phones and ultrabooks. So rather than sticking with one box for 6 years instead they keep two in the air at a given time, a regular model and an "S" revised model. So you launch "xb1" then "xb1 s" in 3 years, then "xb2" 3 years after that, then "xb2 s" 3 years after that, etc, with only the most current two being manufactured and all fully backward compatible. 6 years is too long to sit out hardware changes in my mind, but 3 years isn't bad.

The consoles have an order of magnitude more power headroom to work with, and they have dedicated more silicon up-front. It would take multiple node transitions to match the transistor count, and with the current trends power efficiency would be nowhere near good enough to make up for the higher TDP.

If they do find a way around that problem, the problem becomes the console hardware model. With all the talk about cloud gaming and making the hardware irrelevant, it seems to already be one.
The mobile platforms will have the volumes to justify more frequent hardware transitions, or likely the smaller number of future tablet hardware platforms will have the hundreds of millions of units annually versus tens of millions of consoles over years.

I honestly think we will get to a point where it won't matter, because we will get to a point where an order of magnitude of more hardware isn't enough to really see a big visual difference. In other words I fully expect at some point that we will be able to show a game running on say an ultrabook and running on console that has an order of magnitude more power and there won't be enough visual difference there to make people think it's a big jump. Heck I'd argue we are already there for a percentage of the populace.

In any case I'm saying that they should toss out the old rigid console release model and adapt to what's happening in the marketplace and be ready to release a revised model if needed. What Apple does works well for them, they aren't always tossing everything out and starting new, they release revised "S" models that are cheaper to make but keep iPhones in the media, keeps things interesting an fresh even if they are just a hardware revision, and it lets them keep up to date hardware wise with their competition.

 

[3dilettante]

It would be a continuation of the same platform, so all marketing related costs carry over.

Is the Xbox 2 going to look exactly like the Xbox One, go I the exact same box, and they'll freely mix the two SKUs on the same store shelves?
Is it going to politely not make the Xbox One obsolete and not try to up the ante on the PS4/5?
There's marketing costs incurred when they make a slim, much less a new console.

It would be fully backward compatible so software costs mostly carry over.

The Xbox platform would need to be forward compatible to an unusual degree for that to be true. The older console would need to be able to run everything that can run on the newer one, and to some level of equivalence at that. However, the new console will need to be distinguishable in some aspect like performance or features from the older one, otherwise why bother?

It would be using new hardware funded by Amd because Amd has to be doing that anyways, so no need to spend another billion dollars, just go with Amd's latest and greatest.

AMD would be glad to provide a new semi-custom APU if Microsoft pays for it, although cutting into the long tail of revenue that comes in as Durango or its shrinks mature will probably be factored into that contract.

Many current tablets cost mucho dinero because they are still custom designs based around Arm, hence why they have more costs.

I wouldn't say a customized AMD chip is cheaper. It has advantages in graphics IP and relatively decent CPU cores. The unverified estimated costs for the APUs are much higher than the ARM SoCs tablets and smartphones use.

So like Apple starts with Arm then spends a chunk of time and money to customize things to their needs.

They can do this because in six months more iPhone5s are made than there have ever been Xbox 360s, so the initial costs are spread out over much more hardware.

At the 6 year time frame maybe they need to go more hardcore and spend some money for a new design, but for a 3 year revision model I don't think they need to bother.

Where does Durango Slim fit in this? That's right where a slim model would be introduced for capturing the consumers that aren't on leading edge or in Durango's price bracket.

I honestly think we will get to a point where it won't matter, because we will get to a point where an order of magnitude of more hardware isn't enough to really see a big visual difference. In other words I fully expect at some point that we will be able to show a game running on say an ultrabook and running on console that has an order of magnitude more power and there won't be enough visual difference there to make people think it's a big jump.

Do you expect tablets to be at visual parity with a PS5 or XBox Two in the next 6-8 years?
I'm not confident in their prospects for matching Durango or Orbis in the next three years.
They'll be lucky if they get one node shrink done, or possibly to the hybrid node past 20nm that is a smidgen smaller because it's not a full shrink outside of marketing brochures.

In any case I'm saying that they should toss out the old rigid console release model and adapt to what's happening in the marketplace and be ready to release a revised model if needed.

They probably didn't pick the right architecture or development model if they wanted that. The consoles do push hardware in ways that current ARM systems do not, and the consoles lack a comfortable performance cushion where adding or removing wrinkles to the platform are unlikely to cause problems.
Apple is lucky in that ARM SoCs are traditionally in some kind of sub-basement when it comes to performance, so the expectations aren't there. The mobile chips have started clearing the incredibly low bar they set for themselves, and there's not much good news coming when it comes to 20nm or hybrid node transitions.
If it's not competitive at 28nm from a performance standpoint, it will not have a good chance in the next two nodes of catching up.

 

[joker454]

Is the Xbox 2 going to look exactly like the Xbox One, go I the exact same box, and they'll freely mix the two SKUs on the same store shelves?
Is it going to politely not make the Xbox One obsolete and not try to up the ante on the PS4/5?
There's marketing costs incurred when they make a slim, much less a new console.

They would both be on sale right next to each other on the store shelves exactly how it is right now where 360's and xb1's are both on the same store shelves and being sold side by side. The only difference is that the revised xb1 would be backward compatible, whereas now the two boxes that are on store shelves are not compatible with each other at all. If they can successfully sell two completely incompatible boxes on the same store shelves then I'd argue that they can just as easily sell two boxes that are compatible with each other, just have them at different price points exactly as we see now with the 360 and xb1. As a bonus, xb1 customers would see a future upgrade path rather than having to throw everything away.

The Xbox platform would need to be forward compatible to an unusual degree for that to be true. The older console would need to be able to run everything that can run on the newer one, and to some level of equivalence at that. However, the new console will need to be distinguishable in some aspect like performance or features from the older one, otherwise why bother?

Of course, modern render engines make this quite trivial to support. Simply enable more visual features on the better sku, that's it. They have to support forward compatibility if they want to be taken seriously anyways, the era of forcing users to throw away all their digital content is rapidly coming to and end.

They can do this because in six months more iPhone5s are made than there have ever been Xbox 360s, so the initial costs are spread out over much more hardware.

Sure, that's why I'm not suggesting they go that route and instead go for a cheaper non custom upgrade.

TWhere does Durango Slim fit in this? That's right where a slim model would be introduced for capturing the consumers that aren't on leading edge or in Durango's price bracket.

Nothing changes, they release an xb1 slim as always at a cheaper price point, and release a revised more powerful "xb1 s".

Do you expect tablets to be at visual parity with a PS5 or XBox Two in the next 6-8 years?
I'm not confident in their prospects for matching Durango or Orbis in the next three years.
They'll be lucky if they get one node shrink done, or possibly to the hybrid node past 20nm that is a smidgen smaller because it's not a full shrink outside of marketing brochures.

This is where it gets more nebulous as you have to define visual parity. For example, you can take the 1920x1080 60fps console games and drop them to 1280x720 30fps, that would remove the order of magnitude performance advantage right there and I'd wager most people would consder the visuals to be "almost as good" and not care. It comes down to what you think most people would notice and/or care about. There's not much point for us getting into it here, I'll simply argue that for a large portion of the population providing the same visuals at lower resolution and frame rate would be considered "almost as good" to where they would consider their tablet or ultrabook to be at "visual parity" with consoles.

They probably didn't pick the right architecture or development model if they wanted that. The consoles do push hardware in ways that current ARM systems do not, and the consoles lack a comfortable performance cushion where adding or removing wrinkles to the platform are unlikely to cause problems.
Apple is lucky in that ARM SoCs are traditionally in some kind of sub-basement when it comes to performance, so the expectations aren't there. The mobile chips have started clearing the incredibly low bar they set for themselves, and there's not much good news coming when it comes to 20nm or hybrid node transitions.
If it's not competitive at 28nm from a performance standpoint, it will not have a good chance in the next two nodes of catching up.

I don't really follow why you think their architecture choice and development model aren't ideal. Arm's weren't fast enough at the time of design, and Intel cost too much hence Amd is just about the only choice that was available. As long as they keep a layer between code and the hardware they can offer forward compatibility with relative ease, and being x86 means that even is Amd drops the ball there is still always the possibility of working with Intel one day if anything perhaps for higher end models. Just like now where you can get low end tablets based on Arm and high end tablets based on Intel x86, they could do the same with future consoles, cheaper Amd based models and more expensive Intel based models.

 

[3dilettante]

They would both be on sale right next to each other on the store shelves exactly how it is right now where 360's and xb1's are both on the same store shelves and being sold side by side.

They aren't running massive ad campaigns right now for the Xbox 360, so while I'm not disputing that it is straightforward putting two boxes on the same shelf, the marketing apparatus has to put a little bit more work into it, or so the WiiU has taught me.

If they can successfully sell two completely incompatible boxes on the same store shelves then I'd argue that they can just as easily sell two boxes that are compatible with each other, just have them at different price points exactly as we see now with the 360 and xb1. As a bonus, xb1 customers would see a future upgrade path rather than having to throw everything away.

The Xbox One and 360 are separated by eight years. The 360 had its peak period and is in the declining part of its life cycle.

Creating a new high-end box at that point will mean that the slim is further devalued, while Microsoft's top line would be upgraded hardware that pushes it into the same economic and thermal realm as Durango is right now.

Sure, that's why I'm not suggesting they go that route and instead go for a cheaper non custom upgrade.

The Xbox One is a semi-custom solution. That custom part is going to get in the way of buying a standard part, assuming that AMD is interested in creating a standard desktop part in the next few nodes that can exceed Durango.
I am assuming the new Xbox One successor wouldn't have glaring areas where it regressed relative to the slim, which will probably be the case unless AMD is contracted for a semi-custom design again.

This is where it gets more nebulous as you have to define visual parity. For example, you can take the 1920x1080 60fps console games and drop them to 1280x720 30fps, that would remove the order of magnitude performance advantage right there and I'd wager most people would consder the visuals to be "almost as good" and not care. It comes down to what you think most people would notice and/or care about. There's not much point for us getting into it here, I'll simply argue that for a large portion of the population providing the same visuals at lower resolution and frame rate would be considered "almost as good" to where they would consider their tablet or ultrabook to be at "visual parity" with consoles.

If this is the scenario, the enhanced Durango already sounds pointless, and possibly the Xbox One is the last Microsoft console.
The best that could be hoped for in the next 3-4 years is a possible maximum 30-40% improvement in graphics capability, uncertain CPU gains, and poorer economics for the hardware.
That's just about where the single node transition and three years will get us.

If tablets are supposed to reach some very broad definition of parity with Durango before the end of that console generation, then I don't see how a few tens of percent are going to edge Microsoft over the threshold.
Those same physical challenges will also be afflicting the tablets, so I'm not really counting on them getting to parity that quickly.

I don't really follow why you think their architecture choice and development model aren't ideal. Arm's weren't fast enough at the time of design, and Intel cost too much hence Amd is just about the only choice that was available.

AMD being the best choice for a traditional console doesn't make it the best choice for if Microsoft then decides too change the product life cycle--which went into some of the criteria used to select AMD.
Microsoft doesn't own the IP and does not have license to implement the chip with anyone other than AMD. Any newly contracted designs will mean more checks cut to AMD.

As long as they keep a layer between code and the hardware they can offer forward compatibility with relative ease, and being x86 means that even is Amd drops the ball there is still always the possibility of working with Intel one day if anything perhaps for higher end models.

Three years isn't enough time to contract with Intel for an 8 core CPU design and GCN-level shaders and eSRAM.
No current or near-future low-power x86 core from Intel matches Jaguar's ISA support, and Intel and AMD don't coordinate their architectural weaknesses, so that's not really as compatible as it should be.

I don't see the hardware able to get over the software layer thick enough to abstract away those differences being viable until more nodes go by, which is going to give the Xbox One most of its life cycle uncontested.

 

[Cjail]

PlayStation 4 UK launch sales are over 250k units.

 

[Arwin]

PlayStation 4 UK launch sales are over 250k units.

Good start! So total EU sales may well get close to 1 million then.

 

[loekf]

Good start! So total EU sales may well get close to 1 million then.

There's a rumour floating around:

On the PS4s second release day [Europe, SA, Africa, Asia excluding Japan] its said to have sold around 2.4Million, though that is not yet official.

See: http://www.theinquirer.net/inquirer/news/2316344/sony-s-ps4-is-the-fastest-seller-since-the-psp

Please ignore the comment(s) of Xbone fanboys..

 

[BoardBonobo]

PS4 becomes the UK's fastest selling console in history. Good Job Sony. Or should I say good job MS

Also there is a stock tracker live here UK or US.

Sony are supposed to be setting one up as well.

 

[mrcorbo]

PS4 becomes the UK's fastest selling console in history. Good Job Sony. Or should I say good job MS

What does MS have to do with how many PS4s Sony is able to make?

 

[patsu]

There's a rumour floating around:



See: http://www.theinquirer.net/inquirer/news/2316344/sony-s-ps4-is-the-fastest-seller-since-the-psp

Please ignore the comment(s) of Xbone fanboys..

Hmm... That Inquirer statement seems incorrect. PS4 is not launched in Asia yet right ? I remember reading that South East Asia will launch in mid December. Has this changed ?

http://ap.ign.com/en/news/9141/playstation-4-asia-launch-date-announced

 

[Strange]

Hmm... That Inquirer statement seems incorrect. PS4 is not launched in Asia yet right ? I remember reading that South East Asia will launch in mid December. Has this changed ?

http://ap.ign.com/en/news/9141/playstation-4-asia-launch-date-announced

No it has not. December 18 in Taiwan.

 

[joker454]

They aren't running massive ad campaigns right now for the Xbox 360, so while I'm not disputing that it is straightforward putting two boxes on the same shelf, the marketing apparatus has to put a little bit more work into it, or so the WiiU has taught me.

Well they wouldn't because the 360 is basically end of line since it's not compatible with the xb1. If the two machines available were compatible then marketing would be targeting both at the same time at their respective price points.

The Xbox One and 360 are separated by eight years. The 360 had its peak period and is in the declining part of its life cycle.

It had to be that ridiculously long because they were still stuck on the old console model. In a multi console model there is no need to ever go that long without releasing something new.

Creating a new high-end box at that point will mean that the slim is further devalued, while Microsoft's top line would be upgraded hardware that pushes it into the same economic and thermal realm as Durango is right now.

I don't agree, anymore than NVidia releasing new gpu's devalues their old ones. You can still buy NVidia 6 series cards because there is still demand for them even though Titan's share the same shelf space.

The Xbox One is a semi-custom solution. That custom part is going to get in the way of buying a standard part, assuming that AMD is interested in creating a standard desktop part in the next few nodes that can exceed Durango.
I am assuming the new Xbox One successor wouldn't have glaring areas where it regressed relative to the slim, which will probably be the case unless AMD is contracted for a semi-custom design again.

When Amd has < 28nm process available to them, they can make an apu the same as durango's but with more of the same compute units. That would not be costly to make, it's just a tweak of the existing design taking advantage of the node drop. I'm not thinking of them making entirely new designs every 3 years, think of it as a 3 year tick/tock type schedule like Intel does where Microsoft make an entirely new design at year 0, then refresh it at year 3, then go all new again at year 6.

If this is the scenario, the enhanced Durango already sounds pointless, and possibly the Xbox One is the last Microsoft console.
The best that could be hoped for in the next 3-4 years is a possible maximum 30-40% improvement in graphics capability, uncertain CPU gains, and poorer economics for the hardware.
That's just about where the single node transition and three years will get us.

It's not pointless because there always will be certain people that will buy the new product just for the sake of upgrading, or some people for whom even a 15% speed increase is worth selling the old and buying the new. People do it with gpu's all the time, and since they can sell the box at a profit from day one it makes this feasible to do. Also it lets them target both high and low price points. Right now with one product for 7 years they are stuck in a bad position with pricing. They have to lose low end customers because they can't price it too low, but they can't price it too high because it's not really that strong of a product. They have to land at some compromise price to make it all work. They can toss all that in the garbage with a 2 model system, one can be targeted at the low end customer and one at the high end customer all the time and because they are compatible they can advertise that the users software investment is protected when they decide to upgrade in the future. They can get volume with the low end people and better margin with the high end people. It's how every other product out there works basically, except for consoles. You don't see for example just one NVidua gpu out on the market at a time, or one Apple phone, or one Dell ultrabook, etc, there are always tiers.

If tablets are supposed to reach some very broad definition of parity with Durango before the end of that console generation, then I don't see how a few tens of percent are going to edge Microsoft over the threshold.
Those same physical challenges will also be afflicting the tablets, so I'm not really counting on them getting to parity that quickly.

Tablets will reach "visual parity" for a certain percentage of the buying public. That becomes a problem because then to Microsoft those potential console customers will be permanently lost. The only way to mitigate that is with more frequent hardware refreshes if anything to give the illusion of "new and improved" rather that "7 years old". With new product, new hardware even if just slightly improved and new marketing they can still get back some of these customers. Just like some of them will buy a new iPhone even if it isn't really better than their current one, they just buy it to have the new thing and/or they believe it's actually new and improved to a large extent. They won't go looking at console benchmarks, they will simply buy because it's the new hotness.

AMD being the best choice for a traditional console doesn't make it the best choice for if Microsoft then decides too change the product life cycle--which went into some of the criteria used to select AMD.
Microsoft doesn't own the IP and does not have license to implement the chip with anyone other than AMD. Any newly contracted designs will mean more checks cut to AMD.

Three years isn't enough time to contract with Intel for an 8 core CPU design and GCN-level shaders and eSRAM.
No current or near-future low-power x86 core from Intel matches Jaguar's ISA support, and Intel and AMD don't coordinate their architectural weaknesses, so that's not really as compatible as it should be.

That's true but the high end future in Arm simply is not on the radar right now, at least nothing anywhere near the league of what x86 can offer. Plus there is some confort and safety being back in the x86 camp both fro ma mature tools point of view and knowing that every coder on the planet knows x86. We know that Intel and Amd can be made compatible for the simple reason that it already happens everyday. That's a software problem and one that can clearly be solved, much more easily in this case since it's one particular Amd hardware revision that they would have to emulate on Intel if they went that route. Given what amateur coders have been to emulate on x86 with all the various esoteric console emulators out there I don't really worry about emulating an Amd x86 box on Intel, even less so in this case given the software abstraction layer present on the xb1.

I don't see the hardware able to get over the software layer thick enough to abstract away those differences being viable until more nodes go by, which is going to give the Xbox One most of its life cycle uncontested.

We can check back in 3 years to see where hardware is to determine if there would have been a big enough jump to merit a second sku. I'd ask people here as a test, if in 3 years they could sell their xb1 and get an "xb1 s" that for the sake of argument did nothing more than play every 30fps xb1 game at 60fps, would people here sell their xb1 and buy an "xb1 s"?

 

[Squilliam]

Well their new "xb1" box would be an extension of the old one basically, so you could argue that the same engineering resources spent on the previous one can mostly carry over to the next one. It's not like there is esoteric hardware that cost billions to make which they have to throw away and start over, it's all mostly bog standard stuff. The xb1 already sells at a profit or break even so it's not like the past where they had to wait many years to just recoup launch costs. The new one could use similar tools and hardware, just beefed up versions. I almost think they may have to do this because a tablet running at two manufacturing nodes ahead of a console will start to become serious competition. I just can't imagine them lasting 7 to 8 years on the new console hardware, it's too outdated to last that long.

Being able to never have to do another console launch would probably be a lot less stressful for the hardware manufacturers. It doesn't make sense to constantly have to start from zero especially as the costs of games to take advantage of new hardware is always on the rise. It also makes sense from the perspective that if you're going to have to redesign the chip somewhat to take advantage of a new process node you may as well throw in a little extra performance at the same time.

If they're selling the physical hardware boxes profitably then they don't lose anything by releasing upgraded hardware. There are even fringe benefits to this strategy because as people shift their obsoleted hardware through secondary resale markets they can access more price conscious customers without having to always drop the price of the hardware to rock bottom.

This is a recurring topic on Beyond3d, and I would say that it could be due to the fact that it does have a lot of merits as a strategy. Releasing a newish design in a smaller (cheaper) form factor would certainly shake the market up. I would say that if one console manufacturer did an upgrade in say 2016/17 then it would be an opportunity as well to tempt customers who had gone with the competition to switch as they can offer clear benefits over the existing hardware. Until the iPhone came out and proved that the frequent upgrade cycles worked you could make arguments against this kind of model, but now the pendulum seems to have swung the other way towards a more flexible approach.

 

[blakjedi]

by moving to an already high volume, rapidly advancing architecture (x86) which they intrinsically understand maybe better than anyone else in the world, as well as having relationships with hardware OEMs Xbox can move from a device to another Wintel platform being sold everywhere as a closed box system. Direct X just becomes the minimum spec essentially.

That way you can let the market dictate peoples price points. if I want a 16 CPU quad slip setup... it will run Ryse at a certain level. if I just want a $200 system because 1080p/60 is not that big a deal to me I can purchase that and appreciate my choice. Just like tablets and phones. currently the only wrinkle is Kinect but hopefully that will rapidly drop in price to the point where it becomes a nominal cost.

 

[patsu]

Wintel might be tricky playground because of Windows PC monopoly. May be the ARM platform is easier to start ? Or position it away from Windows ?

 

[wco81]

PC sales are declining.

PC gaming is a smaller market than the console market. If you try to get on a PC upgrade cycle, only a small fraction of the base will upgrade every year.

It must be tempting, when they see hundreds of millions of smart phones and tablets sold every year, at ASPs much greater than consoles, to tap into that kind of dynamic market.

But they may have difficulties matching the sales volume of the last generation, never mind trying to compete against mobile device volumes.

Most of us on forums like this would have no problems buying new console hardware every other year. But that will amount to a few million at most.

If they can build a business model around shipping say 3 million units, at $400 every two years, thats something to consider. However, they aren't going to bail on the current business model of fixed hardware for 5-7 years with $60 games until it's proven that the model is not sustainable.

We may know in 2-3 years whether consoles can continue the way they always have, or whether the onslaught of mobile devices providing "good enough" gaming experiences starts to fritter away or limit the size of the console market.

 

[Dave Baumann]

There's plenty of evidence to say that PC gaming is growing. PC sales are in decline as a whole, but a lot of that is in the commercial sector and in the consumer sector its driven by casuals moving from notebook to tablets. Desktop PC's have remained somewhat more steady and research shows ASP's are going up for gaming hardware:

http://techreport.com/news/25197/as-overall-pc-market-declines-growth-expected-in-gaming-hardware

Likewise, EA for example has seen significant growth in PC games sales:

http://techreport.com/news/25136/ea-financials-show-growing-pc-game-revenue

 

[3dilettante]

It had to be that ridiculously long because they were still stuck on the old console model. In a multi console model there is no need to ever go that long without releasing something new.

The 360's generation lasted as long as it did because the economy tanked and the expected benefits from silicon scaling did not materialize.
The economy is a question mark, but the silicon scaling is mostly expected to be less than what the 360 got.
I'm not sure where the means to manufacture new designs at a doubled cadence is supposed to come from when the processess that make them are slowing dramatically.

I don't agree, anymore than NVidia releasing new gpu's devalues their old ones. You can still buy NVidia 6 series cards because there is still demand for them even though Titan's share the same shelf space.

That required putting a massive professional GPU that gets most of its revenue elsewhere into a higher gaming price bracket typically reserved for dual-GPU cards, and with comparatively small volumes.

Its applicability here is tenuous unless Microsoft does something about the more compressed pricing tiers and the lack of a separate market that GK110 derives most of its revenue from.
The population of gamers that are so price-insensitive that they can foot the bill for a graphics card that is priced double a whole console is vanishingly small. Not even Titan can justify its existence solely on the highest rollers of the gaming segment.

When Amd has < 28nm process available to them, they can make an apu the same as durango's but with more of the same compute units. That would not be costly to make, it's just a tweak of the existing design taking advantage of the node drop.

I don't see where the "just" part comes in.
A "pure" shrink would already be non-trivial. A changed design that bloats the die size back up on a new and more expensive node produces higher costs on top of a new design.
By most projections, 20nm is going to be more expensive for longer, and it's not going to be a major performance upgrade due to it remaining a planar process. For a slim, the less than stellar performance gain isn't important.
The hybrid 16/14nm node will bring FinFETs that will finally produce the power benefits of a traditional node transition, but that will be not be offered for a year or more after 20nm and will not bring significant density over 20nm.
The next two nodes effectively stretch out a node's worth of improvement over two nodes and an additional 1-2 years that doesn't include the customary delay between foundry promises and manufacturing reality.
If we time things relative to when the first AMD GPU at 20nm comes out and assume a 28nm-like curve, 20nm won't look good for a console chip that keeps Durango's die size until late 2015 to mid to late 2016, and the cost projections do not look like they will scale the same way.
If you want the necessary power scaling for a more performant APU, add the time frame for when the hybrid node comes out.

I'm not sure the slim will make full sense in three years, and it's a lower bar to hit than a new APU.

With new product, new hardware even if just slightly improved and new marketing they can still get back some of these customers. Just like some of them will buy a new iPhone even if it isn't really better than their current one, they just buy it to have the new thing and/or they believe it's actually new and improved to a large extent.

The iPhones have advanced performance more than that over the generations. It was primarily helped by the very low bar they had to get over, and Apple had to bite the bullet with increasing amounts of engineering effort and chip customization to help maintain its growth.

We know that Intel and Amd can be made compatible for the simple reason that it already happens everyday. That's a software problem and one that can clearly be solved, much more easily in this case since it's one particular Amd hardware revision that they would have to emulate on Intel if they went that route.

It takes several years to change that sort of thing if such a core is not already in progress, unless you think Microsoft is going to foot the bill for a 4-8 core Broadwell instead of Silvermont.
It is too late to do what you want for Durango Redux unless Microsoft has started this project while trying to finalize Durango.

I'm leaving out the whole eSRAM and GPU question, as how much Intel can catch up and really maintain full compatibility and how much of the implementation information AMD can withold to make it difficult are murky.

 

[Pixel]

With the revelation people getting banned for excessive swearing on Skype and Upload Studios what impact does that have on Microsofts push to get Xbox One and Skype in the business sector for conference calls where trade secrets/pending patents/1st to market ideas/keeping competition in the blue/etc is involved. MS can promise that only computers monitor the data traveling through their skype datacenters, but in the rough and tough jungle of the competitive world where corps stab each other in the back all the time, all the countless internet revalations in the wake of Snowden, how does MS prevent these concerns from dissuading corporate use of Xbox One Skype for conference calls?

Doing a google search just turned out that MS monitors even encrypted skype messages.
http://arstechnica.com/security/201...ssages-get-end-to-end-encryption-think-again/

 

[wco81]

I relied on Skype because of the encryption …

So it's no longer P2P, everything routes through MS servers?