PlayStation 4 Reportedly Codenamed Orbis, Will Feature AMD Hardware, Lock Out Used Games
Rumors about next-generation consoles like the Xbox 720 (codenamed Durango) and the PlayStation 4 have been milling about for several months, but the Sony side of the equation has just gotten a significant update. New information suggests that the PS4 is codenamed Orbis, will be built around an AMD x86-64 CPU and GPU, and will require all new games to registered with a PSN account.
The codename likely means something to Sony; the orbis.scedevnet.com address resolves correctly whereas ps4.scedevnet.com does not.
Let's tackle the report, starting with the hardware specs. Rumors that AMD would provide the GPU tech behind all three consoles date back to last June. Of the three, only Nintendo has confirmed that the Wii U is based on a DX10-era Radeon 4000 part, but there's been bad blood before between Microsoft and Nvidia concerning console development, so AMD's continued presence in that segment would make sense. AMD winning Sony over is a bit more surprising, but sources have previously indicated that AMD was willing to aggressively compete on price.
The irony of an AMD GPU hat trick is that the company's ability to turn a profit on console GPUs isn't all that great. Still, of the two hardware rumors, this one is easier to buy at face value.
AMD Llano Integrated Fusion APU Die Map
An AMD-Powered PS4? Why It Makes Sense -- And Doesn't
Sony has a history of opting for esoteric CPUs with impressive functionality that are both expensive and difficult to use. Sony contributed a hefty chunk of the estimated $400 million it took to develop the Cell processor at the heart of the PS3. In 2005, Ken Kutaragi, onetime President of Sony Computer Entertainment, remarked that the PS3's price was justified because the company wanted consumers to think "I will work more hours to buy one." Four years later, Sony's ex-CEO, Kaz Hirai, admitted that the Cell Broadband Engine was intentionally problematic to develop for. "We don't provide the 'easy to program for' console that [developers] want, because 'easy to program for' means that anybody will be able to take advantage of pretty much what the hardware can do, so then the question is what do you do for the rest of the nine-and-a-half years?" explained Hirai.
Trying to predict what Sony might do with the PS4 is murky for several reasons. IBM, Toshiba, and Sony originally intended to market Cell as a blade server, cluster computing product, and mainframe solution. IBM did provide one commercial update to the Cell architecture Sony uses -- the 2008 PowerXCell 8i, which vastly improved double-precision floating point performance. IBM publicly terminated Cell development after that, and Toshiba jumped ship on its foundry venture with Sony in 2010.
Opting for anything as conventional as an x86 processor would be a sea change for the company, and using an AMD product is a bit curious at this point. On the one hand, AMD might be able to offer Sony an SoC that combined CPU and GPU function to keep costs down -- but such a product would have sharp constraints on die size and any sort of on-die dedicated cache. Even if we assume that AMD can fix Bulldozer's performance problems and substantially ramp up the performance of its integrated GPUs, the idea of launching a console with an estimated 5-8 year life cycle on even an optimistic SoC 18 months from now is...interesting.
To be perfectly blunt, it's hard to imagine that AMD can deliver a CPU with the sort of longevity Sony has demanded from its previous consoles. Granted, a modern GPU can provide the same sort of FPU muscle that made the original Cell such a power house, but the CPU is still important, and Bulldozer's enormous cache latencies cripple the chip. Piledriver is supposedly an incremental improvement, while Kabini/Kaveri aren't expected until 2013. AMD could theoretically be building something custom for Sony, but everything Rory Read has said has pointed towards delayed process node adoption and maximum IP re-use, not expensive custom work.
This one is a puzzle. It could point to a fundamental change in what Sony considers a console, it could be a hint that AMD has much better hardware in the design stages that it isn't talking about yet -- or it could be an incorrect rumor. We'll let you make the call.
Last, But Not Least:
Finally, there's the question of whether or not the console will include a feature that locks out used game functionality. According to rumors, the console will require that all titles be authenticated with a PSN account and one-time activation code. Try to install the game without a new code, and it'll only function in a limited or trial mode. The exact details would presumably be left to each developer, but the idea is to provide publishers with a single lockout service rather than having each company implement its own.
We recently discussed both sides of the issue, ultimately concluding that attempts to lock out the used game market truly have no precedent in other media and are violations of what's known as the First Sale doctrine. Our conclusion owed nothing to the pair of Gamestop defenders who contacted us -- we spoke with one manager and one shill, who variously bombarded us with insults, diatribes, and hyperbolic generalizations, with one brief bout of actual conversation. We left out the insults and red-faced hilarity when we wrote up that conversation, not because we dared not publish, but because the entire episode had become tiresome. The irony is that we were left no choice but to defend Gamestop's business model, despite the best efforts of a Gamestop manager.
The really ugly problem with this sort of lock out is that it won't be used to keep prices lower -- on the contrary, we've seen game publishers talk about how the used market is theft and forcing prices up, despite plenty of evidence to indicate that game demand fluctuates considerably with price. Games aren't like gasoline -- they aren't a product people buy because they have to get to work and changing one's consumption is extremely difficult (and expensive) after a certain point.
The problems with this approach are legion. Would the price of new authentication codes scale downwards as a game's price drops? A 2010 survey reported that 22% of PS3 owners don't have their console connected to the Internet -- are those buyers SOL? We've spoken out against Gamestop's monopoly on used game prices, and we remain opposed to some of the company's practices, but the idea that publishers and Sony are a trusted alternative is positively laughable.
There's no faster way to drive people towards tablet/phone gaming and freemium games that don't cost anything than to try and kill the used market.
But again, just a rumor. Of the ones gathered here, it's the one we hope most is wrong.
The codename likely means something to Sony; the orbis.scedevnet.com address resolves correctly whereas ps4.scedevnet.com does not.
Let's tackle the report, starting with the hardware specs. Rumors that AMD would provide the GPU tech behind all three consoles date back to last June. Of the three, only Nintendo has confirmed that the Wii U is based on a DX10-era Radeon 4000 part, but there's been bad blood before between Microsoft and Nvidia concerning console development, so AMD's continued presence in that segment would make sense. AMD winning Sony over is a bit more surprising, but sources have previously indicated that AMD was willing to aggressively compete on price.
The irony of an AMD GPU hat trick is that the company's ability to turn a profit on console GPUs isn't all that great. Still, of the two hardware rumors, this one is easier to buy at face value.
AMD Llano Integrated Fusion APU Die Map
An AMD-Powered PS4? Why It Makes Sense -- And Doesn't
Sony has a history of opting for esoteric CPUs with impressive functionality that are both expensive and difficult to use. Sony contributed a hefty chunk of the estimated $400 million it took to develop the Cell processor at the heart of the PS3. In 2005, Ken Kutaragi, onetime President of Sony Computer Entertainment, remarked that the PS3's price was justified because the company wanted consumers to think "I will work more hours to buy one." Four years later, Sony's ex-CEO, Kaz Hirai, admitted that the Cell Broadband Engine was intentionally problematic to develop for. "We don't provide the 'easy to program for' console that [developers] want, because 'easy to program for' means that anybody will be able to take advantage of pretty much what the hardware can do, so then the question is what do you do for the rest of the nine-and-a-half years?" explained Hirai.
Trying to predict what Sony might do with the PS4 is murky for several reasons. IBM, Toshiba, and Sony originally intended to market Cell as a blade server, cluster computing product, and mainframe solution. IBM did provide one commercial update to the Cell architecture Sony uses -- the 2008 PowerXCell 8i, which vastly improved double-precision floating point performance. IBM publicly terminated Cell development after that, and Toshiba jumped ship on its foundry venture with Sony in 2010.
Opting for anything as conventional as an x86 processor would be a sea change for the company, and using an AMD product is a bit curious at this point. On the one hand, AMD might be able to offer Sony an SoC that combined CPU and GPU function to keep costs down -- but such a product would have sharp constraints on die size and any sort of on-die dedicated cache. Even if we assume that AMD can fix Bulldozer's performance problems and substantially ramp up the performance of its integrated GPUs, the idea of launching a console with an estimated 5-8 year life cycle on even an optimistic SoC 18 months from now is...interesting.
To be perfectly blunt, it's hard to imagine that AMD can deliver a CPU with the sort of longevity Sony has demanded from its previous consoles. Granted, a modern GPU can provide the same sort of FPU muscle that made the original Cell such a power house, but the CPU is still important, and Bulldozer's enormous cache latencies cripple the chip. Piledriver is supposedly an incremental improvement, while Kabini/Kaveri aren't expected until 2013. AMD could theoretically be building something custom for Sony, but everything Rory Read has said has pointed towards delayed process node adoption and maximum IP re-use, not expensive custom work.
This one is a puzzle. It could point to a fundamental change in what Sony considers a console, it could be a hint that AMD has much better hardware in the design stages that it isn't talking about yet -- or it could be an incorrect rumor. We'll let you make the call.
Last, But Not Least:
Finally, there's the question of whether or not the console will include a feature that locks out used game functionality. According to rumors, the console will require that all titles be authenticated with a PSN account and one-time activation code. Try to install the game without a new code, and it'll only function in a limited or trial mode. The exact details would presumably be left to each developer, but the idea is to provide publishers with a single lockout service rather than having each company implement its own.
We recently discussed both sides of the issue, ultimately concluding that attempts to lock out the used game market truly have no precedent in other media and are violations of what's known as the First Sale doctrine. Our conclusion owed nothing to the pair of Gamestop defenders who contacted us -- we spoke with one manager and one shill, who variously bombarded us with insults, diatribes, and hyperbolic generalizations, with one brief bout of actual conversation. We left out the insults and red-faced hilarity when we wrote up that conversation, not because we dared not publish, but because the entire episode had become tiresome. The irony is that we were left no choice but to defend Gamestop's business model, despite the best efforts of a Gamestop manager.
The really ugly problem with this sort of lock out is that it won't be used to keep prices lower -- on the contrary, we've seen game publishers talk about how the used market is theft and forcing prices up, despite plenty of evidence to indicate that game demand fluctuates considerably with price. Games aren't like gasoline -- they aren't a product people buy because they have to get to work and changing one's consumption is extremely difficult (and expensive) after a certain point.
The problems with this approach are legion. Would the price of new authentication codes scale downwards as a game's price drops? A 2010 survey reported that 22% of PS3 owners don't have their console connected to the Internet -- are those buyers SOL? We've spoken out against Gamestop's monopoly on used game prices, and we remain opposed to some of the company's practices, but the idea that publishers and Sony are a trusted alternative is positively laughable.
There's no faster way to drive people towards tablet/phone gaming and freemium games that don't cost anything than to try and kill the used market.
But again, just a rumor. Of the ones gathered here, it's the one we hope most is wrong.
