Feasibility of an upgradeable or forwards compatible console *spawn*

[dagnastyep]

There have been various rumors floating around that MS would release an "xbox TV" or an "xbox lite" alongside the 720 in 2013.

MS submitted a patent for scalable console hardware in Dec of 2010. If MS releases these two different version of xbox this year, is it feasible that the lower end unit (TV/Lite) could later be upgraded to the higher end (720) unit?

The xbox tv is supposedly a set-top box that would run on the core components of Windows 8, and being able to play casual games. Whereas the previous rumor of the xbox lite says it would support all the current and future entertainment apps, as well as arcade games and Kinect.

I haven't seen or heard many rumors for specs of what this device would use. The SoC from the 360 would work, but probably draw way too much power. One of the rumors points to an ARM SoC, but I wonder if that would offer enough computation power to handle arcade games and kinect. I was thinking what if they used 2-4 jaguar cores with 2-6 GCN CUs in an SoC. That should be relatively lower power, and should be able to handle casual/arcade games.

For consumers, you could buy an xbox lite, or the xbox 720 (which would have the small SoC from the lite as well as a larger SoC for running core games that could be powered on/off as need). Would it make sense if MS made a "720 upgrade module" that you could plug your xbox lite into, and turn it into a full 720? In a way it would be similar with what they did with the HDDVD drive. And the 2 SoCs would communicated using the "communication fabric" detailed in the patent.

I'm not sure if this is feasible or not, but I thought it was an interesting idea and wanted to hear what you guys thought.

 

[Shifty Geezer]

The upgrade hardware idea has been hashed out before, it's still a terrible idea. Consoles make most of their money from software so streamlining development is important.

In the current market and form. Market's evolve and it's premature to say that cannot happen with consoles and they'll be stuck in with a singular market format. A major reason for the software revenue model is because the hardware was released uber-powerful, lossy, and with long legs. Historically it wasn't good to run a fat abstraction layer that slowed the machine down, but with more power, the overhead is proportionally diminishing making forwards compatibility more readily achievable.

I'm quite sure if MS released an XB360+ a year or two ago that played the same games at better framerates, they'd have got enough sales to make money on the hardware, also extending the platform with hand-me-down consoles. If Sony had a PS3+ that could run Borderlands at 60 fps etc., they may have got an extra £50 out of me when I needed a replacement. nVidia have announced a handheld console that'll be getting yearly updates running Android. Steam are wanting a box that'll be playing games that can be run on updated hardware even if they don't bring out a Steambox 2 (which I imagine they would). And the BC thread shows plenty of people value forwards compatibility and the chance to run their existing library in better quality, so a forwards compatible architecture is certainly a valid consideration and that means the potential for more frequent refreshes.

 

[AlphaWolf]

All these platforms and not one big budget AAA game for them. I think it would be terrifying trying to release a huge AAA game that targets hardware that might go through 3 or more iterations during the games production.

 

[JasonLD]

The consoles having yearly updates (like smartphones and tablets) could be very interesting idea. The single biggest flaw in the console business is that the market share resets in every console generation, and yearly updates might eliminate all that.

 

[rekator]

Subscription model with 3 years( more a less one) renew hardware.

 

[Shifty Geezer]

The consoles having yearly updates (like smartphones and tablets) could be very interesting idea. The single biggest flaw in the console business is that the market share resets in every console generation, and yearly updates might eliminate all that.

Every year seems unlikely. But every two or three is very possible IMO.

 

[Squilliam]

Every year seems unlikely. But every two or three is very possible IMO.

What about extending the hardware with cloud based rendering? I always wanted to explore the possibility of combining local and cloud computing on the same tasks. I was reading through wikipedia about algorithm efficiency and being able to precompute some things would be an excellent way to use gobs of memory and certainly would make up for some of the deficit in terms of pure flops.

 

[Nesh]

Every year seems unlikely. But every two or three is very possible IMO.

I cant help it but see a decline of consoles if this begins to be applied as it eliminates one of its advantages as stable and simple systems to use and measure consumer reaction.

I dont see the casual market and hardware upgrading every one or two years as something that go well together.

Consoles owe a huge part of their success to their simplicity, lasting and foolproof utilization. Little Johnny buys a console and is just concerned with what games to buy not hardware compatibility/performance. Of course you could say if little Johnny doesnt care he simply wont upgrade.

But it doesnt really go like this. Developers always squeeze out as much as possible out of available hardware. If the hardware is upgraded they will want to exploit it in areas other than framerate and better resolution. Hello 30fps games and "X" resolution again on the upgraded version but with better graphics. So in the middle of the product's lifespan they will have to consider optimizing the same project for the upgraded and older version and that optimization wont involve halving a framerate or a resolution. They will have to think about effects, texture quality, polygon counts etc

This will also introduce a form of uncertainty in an otherwise relatively more certain userbase. How many will be upgrading? How many upgraded? Is it worth the effort to exploit fully the potential of the upgraded hardware? Would the casual upgrade? How many core gamers will be upgrading? How many are they?


It brings out questions that are more common in the tablet/PC/Smartphone market. And if in that area consoles begin to be similar to other markets it will initiate a transition that would have otherwise not existed or if existed might have taken a different route.

That transition will bring the question: why not abandon consoles altogether and replace completely with powerful more featured tablets that get upgraded every 2-3 years that reminiscent consoles? You will say thats the route probably the market should evolve into. But actually its an evolution that is hasted and not necessarily the best for the utility and quality we get from such gaming products. I simply dont accept that the market evolves always positively and towards the most effective and efficient route. Some newly introduced elements in a market change the evolution of said market itself when consumers are left with limited choices that corporations decide to make available. In other occasions some unrelated product (such as tablets) becomes so well established that it affects our habits and available time for other products, up to the point that it affects an originally unrelated industry/market (such as gaming consoles) or other areas of our lives (face to face communication, playing outside, being occupied with hobbies that require physical activity and skills) taking away qualities we originally had.

 

[Shifty Geezer]

Within a known series of devices and not an open platform, compatibility isn't much of an issue. Devs target 8 cores at 1.6 GHz and 400 CUs this year, and then 16 cores at 2.2 GHz and 800 CUs in 3 years, and 16 cores at 3.2 GHz and 1200 CUs some years after that. Same core architecture, same functionality, same APIs, same executables, better performance.

The issue with developers optimising becomes moot because they can't afford to any more. Technology is powerful enough that the can sit on fat APIs and still get excellent performance on screen, and precious few developers want to optimise - they just want to make the games. Targeting upgraded hardware revisions won't be any different to how it is on iPads - target the low end with the high-end getting better performance until there's enough high-end userbase that you can execute that new idea you have that the old hardware isn't powerful to do properly.

It's a natural progression of consoles as they've become more complicated and demanding, and technology has advances. We've also had devs mention things like wanting a single hardware platform. That doesn't mean it will happen, but there are no real roadblocks to it other than a fixed mentality that's tied to legacy market, defined as it was by technical and economic constraints which are ever changing.

 

[Nesh]

Within a known series of devices and not an open platform, compatibility isn't much of an issue. Devs target 8 cores at 1.6 GHz and 400 CUs this year, and then 16 cores at 2.2 GHz and 800 CUs in 3 years, and 16 cores at 3.2 GHz and 1200 CUs some years after that. Same core architecture, same functionality, same APIs, same executables, better performance.

The issue with developers optimising becomes moot because they can't afford to any more. Technology is powerful enough that the can sit on fat APIs and still get excellent performance on screen, and precious few developers want to optimise - they just want to make the games. Targeting upgraded hardware revisions won't be any different to how it is on iPads - target the low end with the high-end getting better performance until there's enough high-end userbase that you can execute that new idea you have that the old hardware isn't powerful to do properly.

It's a natural progression of consoles as they've become more complicated and demanding, and technology has advances. We've also had devs mention things like wanting a single hardware platform. That doesn't mean it will happen, but there are no real roadblocks to it other than a fixed mentality that's tied to legacy market, defined as it was by technical and economic constraints which are ever changing.

Yeah but the only good example we can think of is an iPad. Where squeezing every ounce of performance available and pushing the envelope are never targeted. There no multi-million huge projects from publishing/developer giants trying to compete each other like they do on consoles. Developer demands and our gaming demands are different on such platforms unless there is a non-open platform that is not like that and I am not aware of. Thats why currently the iPad can run away with it while the competing android devices come with countless hardware architectures and upgrades.

The developing mentality and gaming mentality on an iPad is completely different. This is not a like for like situation. When and if what you describe will be the actual applied evolution of consoles, it will be the time that maybe consoles have transformed to products that may be so similar to the tablets of the future that they will eventually vanish, tablets will evolve in their place after consoles begin to be approached with less ambition and provide less ambitious projects.

Natural or not, fixed mentality or not, whatever the reasons consoles may evolve to that it is not necessarily a positive evolution of the console gaming experience but it will be the only and best option available

 

[Squilliam]

The only things which really kind of make sense to me with regards to upgrading the console itself is probably features in the game itself.

For instance.

Gen 1: HD 1080P @ 30FPS
Gen 2: HD 1080P @ 60FPS or 30FPS 3D or Local coop.
Gen 3: HD UHD @30 FPS /60FPS per eye 3D 1080P / 120FPS 1080P
Gen 4: Bump Gen 2 back to gen 1 and drop support for Gen 1.

That way developers are still making the same game and the experience isn't degraded for those on the base unit and people have a fair idea of what they will get and how long the system lasts. It would give people the standard 6 year console cycle but also give people an opportunity to upgrade earlier.

 

[dobwal]

It's a natural progression of consoles as they've become more complicated and demanding, and technology has advances. We've also had devs mention things like wanting a single hardware platform. That doesn't mean it will happen, but there are no real roadblocks to it other than a fixed mentality that's tied to legacy market, defined as it was by technical and economic constraints which are ever changing.

It is not a natural progression for console to adopt a similar release cycle to tabs and smartphones. It is becoming a pet peeve of mine to see people continually comparing iphone/ipad to consoles in terms of shortening the upgrade cycle. It is a superficial comparison to say the least.

Console manufacturer use a hardware subsidization model where software is the dominant revenue generator. Consoles spend a good chunk of their life acting as loss leaders. Decreasing the upgrade cycle has at least two major effects on hardware. It limits the amount of loss MS/Sony/Nintendo can initially incur with each new hardware release. It increases the cost of that hardware. Both effects would be due to smaller volume orders on each derivative hardware part and more investment needed to not only shrink the chips but increase the performance of said chips.

Smartphones' and tabs' profit models are the antithesis to what we find in the console market. Apple's revenue and profit is driven by sales of the hardware itself. If you bought 10 iphones and 10 ipads every year without buying a single item from the app store or itunes over the life of the devices you would still be one of Apple's best customers. If you did the same with the PS3 and Vita, you would be costing Sony money and those devices would be a wasted investment for Sony. Apple's model encourages a shorten release cycle, the current console model encourages longer release schedules.

Furthermore going from current model to a Apple like model would change the underlying dynamics. Under a iphone/ipad model $500/$600 in 2006 dollars wouldn't buy the hardware found in the original launch PS3. A good portion of that $500/$600 would have to represent profits for Sony. So Sony would have had to use lower performing parts or stripped out major features to maintain the same retail price. And given its accepted that Sony took a loss on the original PS3s, one could surmise that Sony would have had to sell the PS3 at $800-$1200 to produce a tidy profit and still provide the hardware found in the original PS3. Meaning the performance we got in 2006 might not have been available in a $500/$600 PS3 until 2008 or 2010 with a two year upgrade cycle.

Under a shorter refresh model where hardware actually drives profits, games like Gears or UC might not have been possible until years after the release of the PS3 and 360 instead of a year after console launch.

And it should be noted that while Apple refresh schedule is much more aggressive in terms of time, in terms of volume sold and profit generated not so much so. Apple sold almost 100 million iphones and 45 million ipad over the first 9 months of 2012. The iphone and its 7 generation of devices have generated more than 30 billion dollars in profits. That probably more profit than those generated by the 360, xbox, PS3, PS2, Wii, GC and DC combined.

 

[Shifty Geezer]

It is not a natural progression for console to adopt a similar release cycle to tabs and smartphones. It is becoming a pet peeve of mine to see people continually comparing iphone/ipad to consoles in terms of shortening the upgrade cycle. It is a superficial comparison to say the least.

Except I wasn't likening it to tablets, but technological and market progress without or without the existence of smart devices. Even without iPads, forwards compatibility and fat APIs with lower relative overheads are desirable within a console ecosystem.

Console manufacturer use a hardware subsidization model where software is the dominant revenue generator.

This is the historical argument, but just because that's how things have been done, doesn't mean they must continue to be done that way.

 

[dobwal]

Except I wasn't likening it to tablets, but technological and market progress without or without the existence of smart devices. Even without iPads, forwards compatibility and fat APIs with lower relative overheads are desirable within a console ecosystem.

They are desirable in terms of being a nice feature to have. They aren't desirable in terms of the cost and inflexibility that they present. Nor are they a feature that the market demands. Its those reasons along with the shelf life of consoles that make forward compatibility a non-standard feature in consoles.

This is the historical argument, but just because that's how things have been done, doesn't mean they must continue to be done that way.

Yes, but the current model is a naturally progression away from the typical profit model of most tech products. The current model isn't a product of happen stance or chance. It evolve out the reality where software licensing fees deemphasized hardware profits. Console manufacturers pour money into hardware in an effort to beef up performance which leads to consoles being cost neutral to cost negative but makes console more marketable. The console market is propped up by $60 games while the smartphones market is propped up by $800-$1200 in yearly subscription fees and the tab market is propped up by 30-40% profit margin on hardware. The market conditions of those markets are different from the console market and make a shorter release cycles possible.

Most arguments espousing such a move are based on it being possible not being practical. Tell me how this move will sell more games as game sales as a metric can be used to measures benefits seen by gamers, developers as well as manufacturers?

 

[Shifty Geezer]

They are desirable in terms of being a nice feature to have. They aren't desirable in terms of the cost and inflexibility that they present. Nor are they a feature that the market demands. Its those reasons along with the shelf life of consoles that make forward compatibility a non-standard feature in consoles.

The market doesn't demand - it responds. To date no-one has offered a forwards-compatible console architecture. It's also worth pointing out that there's lots of history outside consoles with forwards compatible tech in computers, many of which were used principally for games.

Most arguments espousing such a move are based on it being possible not being practical. Tell me how this move will sell more games as game sales as a metric can be used to measures benefits seen by gamers, developers as well as manufacturers?

It'll sell the same number of games. I see no reason why it'd sell less. It might even sell more thanks to hand-me-down expansion of the platform market. eg. I buy an XB3, play Borderlands 3 on it at 30 fps. In 3 years the XB4 comes out and I buy it to play Borderlands 3 in 60 fps and give the ol' XB3 to the kids/niece/sister/friend who go out and buy Borderlands 3 GOTY edition to play online with me. The current segregated, long-life platform can't share content that way, and the old platform is so long in the tooth that it lacks active interest (I've known PS1s and 2's get passed down but to people only interested in paying the games it comes with and not buying more themselves).

The only negative to this approach is less impact at launch. You no longer get a monster machine for your money. If your opponent releases a lossy monster machine and you launch a conservative machine, there's good chance you lose market share from the core audience. But if you can hold your own and then release a refresh a few years later at a very reasonable upgrade price, you win back the core, and maintaining forwards compatibility ensures they stay faithful. That's exactly what the PC has - you get an increasingly improving experience over time while keeping your library. Controlled hardware means the same open long-life experience with none of the compatibility issues or technological deterrents.

 

[dobwal]

The market doesn't demand - it responds. To date no-one has offered a forwards-compatible console architecture. It's also worth pointing out that there's lots of history outside consoles with forwards compatible tech in computers, many of which were used principally for games.

Forward compatibility has existed for PC for years and it never was adopted by console market because of the longevity of its product line. Forward compatibility is inconsequential for console because if devs warrant that a older console can support a title its simply releases a port for it. Forward compatibility doesn't encourage adoption of new gen console it encourages continued usage of older consoles. Its one thing for you to expect every new game to support 1-3 year old tech its another to expect it to support 5-7 years old tech.

It'll sell the same number of games. I see no reason why it'd sell less. It might even sell more thanks to hand-me-down expansion of the platform market. eg. I buy an XB3, play Borderlands 3 on it at 30 fps. In 3 years the XB4 comes out and I buy it to play Borderlands 3 in 60 fps and give the ol' XB3 to the kids/niece/sister/friend who go out and buy Borderlands 3 GOTY edition to play online with me. The current segregated, long-life platform can't share content that way, and the old platform is so long in the tooth that it lacks active interest (I've known PS1s and 2's get passed down but to people only interested in paying the games it comes with and not buying more themselves).

The only negative to this approach is less impact at launch. You no longer get a monster machine for your money. If your opponent releases a lossy monster machine and you launch a conservative machine, there's good chance you lose market share from the core audience. But if you can hold your own and then release a refresh a few years later at a very reasonable upgrade price, you win back the core, and maintaining forwards compatibility ensures they stay faithful. That's exactly what the PC has - you get an increasingly improving experience over time while keeping your library. Controlled hardware means the same open long-life experience with none of the compatibility issues or technological deterrents.

So you think that moving to a shorter release schedule which causes the manufacturers and developers to incur more costs is warranted with no increase in game sales? Then what you propose adds no value to the market and it simply there to justify the desire of a small segment of gamers who want more powerful consoles earlier.

In the current model you can simply give your kids/niece/sister/friend your current xbox3 and buy a new one. Why because under the current model your xbox3 will be more powerful than the xbox3 under your proposed model and a new xbox3 will be cheaper under the current model. You practically end up with the same thing at less cost to you. The manufacturers under the current model aren't hamstrung to legacy technology and a shorten life cycle. Thereby able to pass those saving onto the amount of performance they can provide in their console.

And how can you win back a core audience with a upgrade that may be no more powerful than the console they bought three years ago. And what dev would willingly support a console with multiple upgrades that gave up the core audience to a lossy console which has a bigger consolidated userbase. The 360 often gets better third party ports because devs don't want the additional cost of milking extra performance out of the PS3. You think devs are going to bother with maximizing a port for your new console iteration when a significant portion of the userbase is on the beefier but lossy console and the majority of your userbase is still on your previously released model?

You seem to forget that fully supporting the console market requires a dev to produce ports for at least 3 different consoles and 2 different handhelds. The number grows when your releases are close to transition points between current and next gen hardware. There is no forward compatibility and API to handle that. Shortening the release cycle will only muddy the waters.

 

[Bagel seed]

With streaming and cloud games on the horizon it's doubtful any console being upgradeable or forwards compatible really matters in the next 10 years so it's all moot.

But I'll say that this whole upgrade mentality for consoles only really works for 6-8 year cycles where casuals may even begin to notice a difference from the hardware leap. Increases in performance from process node shrinks refreshed parts are supposed to lower the cost to open up the market to the mainstream which is how it's always worked. And this gen's only begun to slow down in sales noticeably this year or so. Not too bad after all this time IMO.

And phones and tablets are more like fashion accessories these days anyways. Dat peer pressure to upgrade! Also they can get away with it because the advancements are still noticeable to the average person. And, it's basically an entirely new market which is at massive hype levels like when Walkman was all the craze in the 80s. Console games are getting that passe feeling.

I wonder if Sony could've gotten away with such a model back in PSX, PS2 days, with new and faster models released every other year. Probably yes, with a subsidized model like cell phones. But like I said before, we'll never find out if it's feasible since the dumb terminal model is coming.

Other barriers to success: increased dev effort, costs to them and complications.

Who it would sell most to: core fans of the platform. But why sell a faster console to them when it's been proven you can sell the same one to them again at a greater profit? ie. PS3/360 slim

 

[SedentaryJourney]

Within a known series of devices and not an open platform, compatibility isn't much of an issue. Devs target 8 cores at 1.6 GHz and 400 CUs this year, and then 16 cores at 2.2 GHz and 800 CUs in 3 years, and 16 cores at 3.2 GHz and 1200 CUs some years after that. Same core architecture, same functionality, same APIs, same executables, better performance.

The issue with developers optimising becomes moot because they can't afford to any more. Technology is powerful enough that the can sit on fat APIs and still get excellent performance on screen, and precious few developers want to optimise - they just want to make the games. Targeting upgraded hardware revisions won't be any different to how it is on iPads - target the low end with the high-end getting better performance until there's enough high-end userbase that you can execute that new idea you have that the old hardware isn't powerful to do properly.

It's a natural progression of consoles as they've become more complicated and demanding, and technology has advances. We've also had devs mention things like wanting a single hardware platform. That doesn't mean it will happen, but there are no real roadblocks to it other than a fixed mentality that's tied to legacy market, defined as it was by technical and economic constraints which are ever changing.

What's the business case for bi-annual or tri-annual upgrades? Since performance optimizations are still being made for the older base level platform, and I doubt this is going away next gen, what benefit does a slightly newer console give to justify the cost of its development? The current model has the price of a console decrease over time, the bi-annual upgrade model doesn't even have this advantage.

To be honest, if you look at the resurgence of PC gaming I doubt much if any of it has to do with the PC upgrade cycle. It's more about the open nature of the platform and relative ease of publishing for services like Steam.

Even then, these are mostly smaller titles we're talking about. AAA development still leads on consoles where the publishers are making back their money. I agree that gaming is changing, but I find tech level comparisons are really just missing the mark. Sony and Microsoft would be better served by working on their licensing/publishing model if they want to draw indies or prevent publishers from abandoning ship.

 

[Shifty Geezer]

What's the business case for bi-annual or tri-annual upgrades? Since performance optimizations are still being made for the older base level platform,

Optimisations won't be made because the games will be running on fat APIs for device compatibility.

and I doubt this is going away next gen, what benefit does a slightly newer console give to justify the cost of its development? The current model has the price of a console decrease over time, the bi-annual upgrade model doesn't even have this advantage.

Yes it does. The old console can be the entry level model; the new tech costs more.

 

[JasonLD]

What's the business case for bi-annual or tri-annual upgrades? Since performance optimizations are still being made for the older base level platform, and I doubt this is going away next gen, what benefit does a slightly newer console give to justify the cost of its development? The current model has the price of a console decrease over time, the bi-annual upgrade model doesn't even have this advantage.

I think the biggest advantage of having shorter hardware refresh cycle while maintaining compatibility is that it could effectively remove the console generation cycle. Instead of rebuilding the market share from zero with every console generation, the company can effectively maintain their marketshare while expanding with each new releases.
In this business model, the console manufacturer may also need to make profit on the hardware to justify the r/d costs. It would mean that consoles would be far more modest in terms of processing power than it would for the traditional model, but shorter refresh cycle would mean that the consoles would stay in line with any technological advancements.

Damn it is hard to make long posts with smartphone lol.