If only it stopped at "non-military" uses. The author of GlovePIE seems to be one of those cause-obsessed fanatics - the site is covered with blurbs decrying Scientology, homeopathy, and global warming, to name a few. Even if you agree with him on these things, he comes off as an asshole, because the GlovePIE site seems to talk more about anything except the actual software.
To download, you basically have to agree to only use "100% Green power". The license agreement includes the following clauses: * Cannot be used with military software - explicitly including being used to play America's Army * Cannot be used in Israel or any Israeli-occupied territory until Israel meets a rather long list of demands. He claims to have put some code in there to disable the software if it's run in Israel. * Cannot be used to cheat at multiplayer games (which yeah, I can understand that one)
And somehow he has the audacity to claim some license terms from the Kinect SDK are "fascist".
They're eliminating the socket, so the motherboard and CPU are a package deal now.
What harm does this do?
Well, it makes it impossible to upgrade the CPU without upgrading the motherboard. That's effectively already the case for Intel processors. They only maintain socket compatibility for one generation. Nehalem boards worked under Westmere, Sandy Bridge boards worked under Ivy Bridge, and so on. They don't really keep compatibility long enough for processors to get significantly more powerful, so there's relatively little incentive to upgrade the CPU while keeping the same motherboard.
Yes, it impacts people who want to upgrade from a low-end processor to a high-end. And repairs - if the CPU gets fried, you can't just replace it, and not the mobo. But I would not consider these particularly common circumstances.
So what's the upside?
A slight decrease in price ($10-20 or so, I would venture). A bit more reliability - no more bent pins. It might cut down on the often-crazy number of slightly different models.
Ultimately, it comes down to a move that's a major inconvenience for a very small number of users, a minor problem for a small group, and a minor gain for the majority. And that's just within the enthusiast community - everyone who would never upgrade their processor is in the same "small-gain-no-loss" boat as the enthusiasts who upgrade both their CPU and mobo every 2-3 years or so.
I've always thought that laws regulating drunk driving based on blood alcohol content were missing the point. It's not driving with alcohol in your blood that causes problems, it's driving while your coordination, reaction time and judgement are impaired. Yes, the former causes the latter, but there's no universal amount - I know some people who can drive just fine after a beer or two, while I don't trust myself after just one.
What police should look at is more general sobriety tests. The stuff they used to do, like "recite the alphabet backwards" or "follow my finger with your eyes". Ideally they should be testing for exactly what should be criminalized - that is, coordination, reaction time and judgement. Not only does this work for all existing drugs, but it should work for basically any drug they ever invent.
Erm, the newest Intel graphics can run Crysis at semi-respectable settings. Considering I ran ZSNES just fine on a Rage 3D, I have no idea where you got the idea Intel graphics can't run a simple SNES emulator.
Remember, Intel's newest graphics ("Intel HD") are a ground-up new design, not at all related to the old ("Intel GMA") integrated graphics.
It's not just a question of past and current games, but future games as well.
Will the next Arkham game support Linux? The next Battlefield? The next Crysis? The next Deus Ex? The next Elder Scrolls? Odds are the answer is "no" to all of them, and I'm only five letters into the alphabet.
More than that, will it run well? It's already rough enough, playing console ports on Windows - having to put up with bad control schemes, limited graphics options, often having to do some fiddling just to make a game work, simply because the developers considered Windows to be a second-class system. And if you think they won't consider Linux a second-class citizen or worse, you must be smoking something good.
There's no single game holding gamers onto Windows. There's no group of games holding us on. There's pretty much every game, ever.
Yes, for some of you, one or two games would suffice to pull you over. But ask yourself - are you a "gamer", or are you a "person who plays games sometimes"? As for me, I'm a gamer. This week alone, I played seven different games. Twenty-two in the past month. For people like me, Linux just won't cut it.
At least, not for a while. If 90+% of the games released on Windows also come out on Linux, over the next few years, it will be a serious contender. Or if there's a good, AAA-quality title that is released as a Linux exclusive, that could push things.
But it is not going to be an overnight process. Linux is only recently beginning to appear attractive to developers. Next you'll have to convince the marketing executives, THEN you can start convincing the gamers.
Plan A, stop global warming before it becomes a problem, failed long ago. Global Warming is now an actual, current problem.
Plan B, stop global warming before it becomes an insurmountable problem, is looking more and more bleak. We can't seem to convince anyone that it's a problem even when it actually is. Maybe Plan B can be salvaged, but we need another fallback plan.
Plan C is "Earth is fucked, what now?". We'll also want to genetically engineer crops and livestock that can live in the new conditions (hotter and drier, to put it simplistically), and start migrating cities that are shortly to be underwater inward. I'm also thinking evacuation should be at least one prong of Plan C - with the last of our fossil fuels, we should be able to get a self-sustaining Mars colony started. Hell, if we could export our excess CO2 to Mars somehow, we could lower our temperatures while increasing Mars's to a more human-friendly level. Pity there's no feasible way to do that.
It could very well be a POWER7-based design. Key word "based". Nobody (well, nobody with two brain cells) is saying they put a full POWER7 chip in the console.
But here's what they could have done:
Take a stock POWER7 chip. Strip it down to 4 cores - or maybe down to triple- or dual-core. Strip out some of the more redundant execution units (decimal float? four floating-point units?). Cut down on the massive cache. Cut out all the multi-socket stuff, the ECC support, trim down the memory controller to what a console needs, and lower the clock speed to keep the heat down.
That probably just cut your performance to under a quarter of a full POWER7, but it probably cut your costs down even more, to something that could actually fit in a console.
Good source there, but I have to take that one more skeptically than I normally would take an Ars article - it claims the Wii U has 1GB of memory, which has been demonstrated to be wrong by early teardowns that count 2GB. I'm definitely not going to discount it completely - it's got an actual source who's working with the hardware, after all - but it might not be completely true, based off early prototype hardware or something, maybe.
Did you see some source I haven't? I've been scouring the net regularly for detailed specs on the Wii U, and as of right now, I can't find any reputable specs for the CPU or GPU.
We do know that it's a POWER-based CPU, almost definitely POWER7, but it could be single-core for all we know (although the rumors seem to have settled on quad-core, with some level of SMT, with a clock speed in the 3GHz range). And the GPU seems to be a complete mystery, other than it being made by AMD.
I'm not saying you're wrong, I'm more curious as to where you got that info so I can read it myself.
I'll also note that, if the rumors are right, it basically confirms my "half-generation" hypothesis, that Nintendo is deliberately designing their consoles to be half a generation behind Microsoft/Sony, so they get lower hardware costs, better thermal bounds, and can just follow the architecture of the "winning" console instead of risking a less established architecture, but are still "close enough" to the current-gen to be competitive for the hardcore gamers, and are enough of an improvement on the last generation to entice their own customers to upgrade.
It's more of "this is how Valve could make a console" than "this is how a console will run Linux".
Valve doesn't really have the manpower or the expertise to develop their own operating system. And the licensing costs to put Windows (or god forbid OS X) on their console would double the price tag, if they could even get the licensing.
But if they port to Linux, it's a short leap software-wise to making their own console.
So you're misunderstanding the story. It's "Steam on Linux + Big Picture Mode = Steam Console (maybe?)". Linux support is just one of the ancillary things.
PS: Valve generally offers developers a far larger cut of Steam revenues than is traditional (rumor has the norm being a 70-30 split, favoring the developer), and they've already spent years building a particularly strong brand identity and relationships with other major publishers. They probably have as much a chance of pulling off a next-gen console as any of the established players.
PPS: This could, however, be good for Linux gaming. As it is, Steam on Linux will only support those games that are ported to Linux, which generally isn't worth the effort. But if there's a successful Linux-based Steam console, it improves the ROI by making the port easier. It's easier to argue to a marketing exec "we've got a game for the SteamBox (50 million users), so porting to Linux is basically just testing it under a few major distros and fixing some packaging things so we can get another hundred thousand users or so." than it is to argue "we can spend three man-months porting the game to Linux and we'll probably see 50,000 sales".
As of this second, three million, two hundred and fifty-four thousand, seven hundred and seventy-three Steam users are online. They've peaked over five million. So yes, a lot of people "still" use Steam.
Always late with patches.
Can you give a citation there? I've never noticed them be particularly late to patch a game - in fact, they seem to do so faster than PSN/XBox Live. It probably does vary quite a bit depending on the game, though.
Their wrapper often breaks games or adds instability.
Another citation, if you would, please? I've only noticed that (rarely) with the Steam Overlay, which is easily disabled (both globally and on a per-game basis). And even then, all it did for me was kick me off some BF2 servers as a "cheat".
Customer service is non-existent.
While I haven't personally ever needed to speak to them, the reputation of Steam's customer service seems to have improved greatly over the years. I know back around 2006 or so they had a horrible reputation, but it's been years since I heard any complaints (a sharp contrast to Origin or Blizzard, in particular).
Yeah no there are plenty of other options for buying/downloading legitimate games online.
And you're welcome to use them. But how many of them are even thinking about Linux support?
Good luck with the linux project. I want nothing to do with Steam.
And you felt the need to shout that out for everyone to hear? Makes me wonder if you ever actually used it.
Those aren't bomb materials. Those are, at best, timer-making materials and a place to conceal something.
Yeah, sure, if the guy's got ammonium nitrate and kerosene in his carry-on, those would be bomb-making materials. Go ahead and arrest him for that.
But if you're going to start arresting people for stuff that can be part of a bomb without actually being dangerous, you should start by taking away everyone's cell phone. Not only do they all have timer functions now, but they can, and have, been used as remote triggers.
It's faster than Flash, but not as high-capacity as DRAM (and it's probably a bit slower than DRAM as well). Just like Flash-based SSDs fit between DRAM and hard drives.
Right now, it uses the interface of DRAM. They could probably have shoved it in a PCIe or even a SATA interface, but it was more logical to use the fastest possible.
Overall, I'm not sure where it's headed. If they can't get density up, it's doomed to niche uses, like embedded hardware. Maybe replace battery-backed RAM in things like RAID controllers. But depending on how they can scale the density (and how performance scales with density), they could replace DRAM, Flash, or maybe even SRAM.
Blizzard/Activision/EA could buy AMD though, make their own consoles.
Not really likely. If you want to break into the console market, you want to buy makers of prebuilt computers, operating systems, input peripherals. You can easily make a console using someone else's chips (I don't think anyone's made one off their own chip design, actually). In-housing the parts is a move an established console maker would do to gain an edge, not something a first-timer would do.
"Microsoft Corp, Google Inc, Samsung Electronics, Intel Corp and even Facebook Inc have been suggested by Wall Street analysts as potential suitors"
Intel would never buy AMD. Face it - right now, Intel is *winning* in the market, pretty much legitimately (not 100%, and they used to cheat like mad, but right now they're winning more-or-less fairly). But they need a competitor to avoid a massive antitrust investigation. They need AMD as an enemy more than they need it as an asset.
Facebook would not, and could not, buy AMD. They may be riding high on the Web 2.0 Bubble, but they're an absolutely terrible match. Facebook's made it a point of using off-the-shelf hardware and open-source solutions. They have very little experience with hardware (besides setting up networks and racks), and gain nothing from producing their own hardware.
Google doesn't need them. They're doing fine running on commodity servers for their web stuff, and trying to produce their own mobile chips would anger their hardware partners for Android. It might give them a slight edge in the long run, but the short-term harm seems to outweigh that.
Microsoft *might* work. They need some special edge in the tablet war they just jumped into, and AMD is a good match with their successful Xbox line. But AMD isn't known to be particularly good at low-power chips. Perhaps they just haven't tried yet, or some older design could be successfully adapted into tablets (a single/dual-core, low-power K8 paired with a good Radeon design might be a good A6 competitor, especially if Microsoft tries to bill itself both as an 'enterprise' tablet *and* a 'gaming' tablet). But really, although it makes sense for Microsoft to buy some hardware company, AMD isn't the best choice. NVidia might make a better one, but I don't think they're looking to sell out right now.
Samsung might buy parts of the company, but they wouldn't want the whole thing. I imagine they would love the graphics section, maybe some of the CPU engineers, but I doubt they want to enter the full-on CPU market.
You know who might make more sense? Cray, or maybe IBM. AMD stuff is popular for supercomputers, both their Opterons and their FireStream/FirePro cards. IBM isn't too likely (they have enough good hardware people already), but Cray or one of their competitors seems at least more plausible than any of the other suggestions.
Another idea is some gaming company. AMD has a somewhat-competitive graphics division, and a compute side that could handle gaming loads well with some tweaks. Sony is really the most likely - they've *never* been good at the hardware side, only lucking into success with the PS1 and PS2 after some clever business decisions. But I also doubt Sony is smart enough to try to do that, especially since buying AMD might hurt their (Intel-focused) laptop business.
Not quite true. Ever since the Pentium Pro (or somewhere around there), there's been something called "Physical Address Extensions", or PAE. This basically just lets the OS give each process a different 32-bit partition of a 48-bit address space.
32-bit Linux systems generally support it. I believe OS X used it as well on 32-bit processors, before they went purely x86-64.
Windows "supports" it, but doesn't have it enabled by default, and I think they block the option on non-server versions of Windows.
Ironically, I don't think you can. At least, not at all easily.
The FirePro cards shows don't support multiple GPUs driving a single output. So you'd have to find some way to avoid treating the array of screens as a single screen, which is far easier said than done. At the very least, extensive reworking of the game's renderer seems in order, and any game open-source enough for a modder to do that is old or graphically-limited enough that you could drive quad QHD monitors using my laptop.
The Tesla and Xeon Phi don't support video output at all, and the drivers have no OpenGL or Direct3D support to my knowledge. So you would have to reimplement the entire graphics layer, then somehow pipe it out through another video card via shared memory.
And honestly, even then, I'm not sure you'd need a array of eight of these cards for any game on the market. I can max out nearly any game on my laptop, at full 1080p, using the mobile version of the GeForce 640. A mere 384 shader cores. The Tesla K20X is based on the same architecture, but scaled up to 7 times as many cores, and at a clock speed not significantly lower. So a gaming-oriented version of the K20X could max out any game on a three-by-two array of "Full HD" monitors (5760x2160). At four, you're running a three-by-two array of "Ultra HD" monitors (11520x4320) at 120+FPS. There is no way I can think of to feasibly require more from your hardware than this, and that's half the GPUs you proposed.
Slashdot Mirror
If only it stopped at "non-military" uses. The author of GlovePIE seems to be one of those cause-obsessed fanatics - the site is covered with blurbs decrying Scientology, homeopathy, and global warming, to name a few. Even if you agree with him on these things, he comes off as an asshole, because the GlovePIE site seems to talk more about anything except the actual software.
To download, you basically have to agree to only use "100% Green power". The license agreement includes the following clauses:
* Cannot be used with military software - explicitly including being used to play America's Army
* Cannot be used in Israel or any Israeli-occupied territory until Israel meets a rather long list of demands. He claims to have put some code in there to disable the software if it's run in Israel.
* Cannot be used to cheat at multiplayer games (which yeah, I can understand that one)
And somehow he has the audacity to claim some license terms from the Kinect SDK are "fascist".
They're eliminating the socket, so the motherboard and CPU are a package deal now.
What harm does this do?
Well, it makes it impossible to upgrade the CPU without upgrading the motherboard. That's effectively already the case for Intel processors. They only maintain socket compatibility for one generation. Nehalem boards worked under Westmere, Sandy Bridge boards worked under Ivy Bridge, and so on. They don't really keep compatibility long enough for processors to get significantly more powerful, so there's relatively little incentive to upgrade the CPU while keeping the same motherboard.
Yes, it impacts people who want to upgrade from a low-end processor to a high-end. And repairs - if the CPU gets fried, you can't just replace it, and not the mobo. But I would not consider these particularly common circumstances.
So what's the upside?
A slight decrease in price ($10-20 or so, I would venture). A bit more reliability - no more bent pins. It might cut down on the often-crazy number of slightly different models.
Ultimately, it comes down to a move that's a major inconvenience for a very small number of users, a minor problem for a small group, and a minor gain for the majority. And that's just within the enthusiast community - everyone who would never upgrade their processor is in the same "small-gain-no-loss" boat as the enthusiasts who upgrade both their CPU and mobo every 2-3 years or so.
I've always thought that laws regulating drunk driving based on blood alcohol content were missing the point. It's not driving with alcohol in your blood that causes problems, it's driving while your coordination, reaction time and judgement are impaired. Yes, the former causes the latter, but there's no universal amount - I know some people who can drive just fine after a beer or two, while I don't trust myself after just one.
What police should look at is more general sobriety tests. The stuff they used to do, like "recite the alphabet backwards" or "follow my finger with your eyes". Ideally they should be testing for exactly what should be criminalized - that is, coordination, reaction time and judgement. Not only does this work for all existing drugs, but it should work for basically any drug they ever invent.
Thanks to the newest data from Curiosity, NASA has finally determined, with 99% confidence, why kids love Cinnamon Toast Crunch.
Erm, the newest Intel graphics can run Crysis at semi-respectable settings. Considering I ran ZSNES just fine on a Rage 3D, I have no idea where you got the idea Intel graphics can't run a simple SNES emulator.
Remember, Intel's newest graphics ("Intel HD") are a ground-up new design, not at all related to the old ("Intel GMA") integrated graphics.
It's not just a question of past and current games, but future games as well.
Will the next Arkham game support Linux? The next Battlefield? The next Crysis? The next Deus Ex? The next Elder Scrolls? Odds are the answer is "no" to all of them, and I'm only five letters into the alphabet.
More than that, will it run well? It's already rough enough, playing console ports on Windows - having to put up with bad control schemes, limited graphics options, often having to do some fiddling just to make a game work, simply because the developers considered Windows to be a second-class system. And if you think they won't consider Linux a second-class citizen or worse, you must be smoking something good.
There's no single game holding gamers onto Windows. There's no group of games holding us on. There's pretty much every game, ever.
Yes, for some of you, one or two games would suffice to pull you over. But ask yourself - are you a "gamer", or are you a "person who plays games sometimes"? As for me, I'm a gamer. This week alone, I played seven different games. Twenty-two in the past month. For people like me, Linux just won't cut it.
At least, not for a while. If 90+% of the games released on Windows also come out on Linux, over the next few years, it will be a serious contender. Or if there's a good, AAA-quality title that is released as a Linux exclusive, that could push things.
But it is not going to be an overnight process. Linux is only recently beginning to appear attractive to developers. Next you'll have to convince the marketing executives, THEN you can start convincing the gamers.
Plan A, stop global warming before it becomes a problem, failed long ago. Global Warming is now an actual, current problem.
Plan B, stop global warming before it becomes an insurmountable problem, is looking more and more bleak. We can't seem to convince anyone that it's a problem even when it actually is. Maybe Plan B can be salvaged, but we need another fallback plan.
Plan C is "Earth is fucked, what now?". We'll also want to genetically engineer crops and livestock that can live in the new conditions (hotter and drier, to put it simplistically), and start migrating cities that are shortly to be underwater inward. I'm also thinking evacuation should be at least one prong of Plan C - with the last of our fossil fuels, we should be able to get a self-sustaining Mars colony started. Hell, if we could export our excess CO2 to Mars somehow, we could lower our temperatures while increasing Mars's to a more human-friendly level. Pity there's no feasible way to do that.
That's a port. Now get an emulator that runs an unmodified ROM of FF3 to do that.
Some things are trivial in a port, but extremely difficult for an emulator.
Ah, that might explain things.
Although that only raises further questions, like "what the hell is the system doing to require a full gig of memory?"
It could very well be a POWER7-based design. Key word "based". Nobody (well, nobody with two brain cells) is saying they put a full POWER7 chip in the console.
But here's what they could have done:
Take a stock POWER7 chip. Strip it down to 4 cores - or maybe down to triple- or dual-core. Strip out some of the more redundant execution units (decimal float? four floating-point units?). Cut down on the massive cache. Cut out all the multi-socket stuff, the ECC support, trim down the memory controller to what a console needs, and lower the clock speed to keep the heat down.
That probably just cut your performance to under a quarter of a full POWER7, but it probably cut your costs down even more, to something that could actually fit in a console.
Good source there, but I have to take that one more skeptically than I normally would take an Ars article - it claims the Wii U has 1GB of memory, which has been demonstrated to be wrong by early teardowns that count 2GB. I'm definitely not going to discount it completely - it's got an actual source who's working with the hardware, after all - but it might not be completely true, based off early prototype hardware or something, maybe.
Did you see some source I haven't? I've been scouring the net regularly for detailed specs on the Wii U, and as of right now, I can't find any reputable specs for the CPU or GPU.
We do know that it's a POWER-based CPU, almost definitely POWER7, but it could be single-core for all we know (although the rumors seem to have settled on quad-core, with some level of SMT, with a clock speed in the 3GHz range). And the GPU seems to be a complete mystery, other than it being made by AMD.
I'm not saying you're wrong, I'm more curious as to where you got that info so I can read it myself.
I'll also note that, if the rumors are right, it basically confirms my "half-generation" hypothesis, that Nintendo is deliberately designing their consoles to be half a generation behind Microsoft/Sony, so they get lower hardware costs, better thermal bounds, and can just follow the architecture of the "winning" console instead of risking a less established architecture, but are still "close enough" to the current-gen to be competitive for the hardcore gamers, and are enough of an improvement on the last generation to entice their own customers to upgrade.
I don't think so. It will probably only run exactly the way it did on the Wii, much like Gamecube games aren't improved by playing on the Wii.
It's possible they might do some upscaling or antialiasing, though. I don't think it's likely, but it's not implausible.
It's more of "this is how Valve could make a console" than "this is how a console will run Linux".
Valve doesn't really have the manpower or the expertise to develop their own operating system. And the licensing costs to put Windows (or god forbid OS X) on their console would double the price tag, if they could even get the licensing.
But if they port to Linux, it's a short leap software-wise to making their own console.
So you're misunderstanding the story. It's "Steam on Linux + Big Picture Mode = Steam Console (maybe?)". Linux support is just one of the ancillary things.
PS: Valve generally offers developers a far larger cut of Steam revenues than is traditional (rumor has the norm being a 70-30 split, favoring the developer), and they've already spent years building a particularly strong brand identity and relationships with other major publishers. They probably have as much a chance of pulling off a next-gen console as any of the established players.
PPS: This could, however, be good for Linux gaming. As it is, Steam on Linux will only support those games that are ported to Linux, which generally isn't worth the effort. But if there's a successful Linux-based Steam console, it improves the ROI by making the port easier. It's easier to argue to a marketing exec "we've got a game for the SteamBox (50 million users), so porting to Linux is basically just testing it under a few major distros and fixing some packaging things so we can get another hundred thousand users or so." than it is to argue "we can spend three man-months porting the game to Linux and we'll probably see 50,000 sales".
People still use Steam?
As of this second, three million, two hundred and fifty-four thousand, seven hundred and seventy-three Steam users are online. They've peaked over five million. So yes, a lot of people "still" use Steam.
Always late with patches.
Can you give a citation there? I've never noticed them be particularly late to patch a game - in fact, they seem to do so faster than PSN/XBox Live. It probably does vary quite a bit depending on the game, though.
Their wrapper often breaks games or adds instability.
Another citation, if you would, please? I've only noticed that (rarely) with the Steam Overlay, which is easily disabled (both globally and on a per-game basis). And even then, all it did for me was kick me off some BF2 servers as a "cheat".
Customer service is non-existent.
While I haven't personally ever needed to speak to them, the reputation of Steam's customer service seems to have improved greatly over the years. I know back around 2006 or so they had a horrible reputation, but it's been years since I heard any complaints (a sharp contrast to Origin or Blizzard, in particular).
Yeah no there are plenty of other options for buying/downloading legitimate games online.
And you're welcome to use them. But how many of them are even thinking about Linux support?
Good luck with the linux project. I want nothing to do with Steam.
And you felt the need to shout that out for everyone to hear? Makes me wonder if you ever actually used it.
Those aren't bomb materials. Those are, at best, timer-making materials and a place to conceal something.
Yeah, sure, if the guy's got ammonium nitrate and kerosene in his carry-on, those would be bomb-making materials. Go ahead and arrest him for that.
But if you're going to start arresting people for stuff that can be part of a bomb without actually being dangerous, you should start by taking away everyone's cell phone. Not only do they all have timer functions now, but they can, and have, been used as remote triggers.
It's a technology that fits between them.
It's faster than Flash, but not as high-capacity as DRAM (and it's probably a bit slower than DRAM as well). Just like Flash-based SSDs fit between DRAM and hard drives.
Right now, it uses the interface of DRAM. They could probably have shoved it in a PCIe or even a SATA interface, but it was more logical to use the fastest possible.
Overall, I'm not sure where it's headed. If they can't get density up, it's doomed to niche uses, like embedded hardware. Maybe replace battery-backed RAM in things like RAID controllers. But depending on how they can scale the density (and how performance scales with density), they could replace DRAM, Flash, or maybe even SRAM.
Blizzard/Activision/EA could buy AMD though, make their own consoles.
Not really likely. If you want to break into the console market, you want to buy makers of prebuilt computers, operating systems, input peripherals. You can easily make a console using someone else's chips (I don't think anyone's made one off their own chip design, actually). In-housing the parts is a move an established console maker would do to gain an edge, not something a first-timer would do.
"Microsoft Corp, Google Inc, Samsung Electronics, Intel Corp and even Facebook Inc have been suggested by Wall Street analysts as potential suitors"
Intel would never buy AMD. Face it - right now, Intel is *winning* in the market, pretty much legitimately (not 100%, and they used to cheat like mad, but right now they're winning more-or-less fairly). But they need a competitor to avoid a massive antitrust investigation. They need AMD as an enemy more than they need it as an asset.
Facebook would not, and could not, buy AMD. They may be riding high on the Web 2.0 Bubble, but they're an absolutely terrible match. Facebook's made it a point of using off-the-shelf hardware and open-source solutions. They have very little experience with hardware (besides setting up networks and racks), and gain nothing from producing their own hardware.
Google doesn't need them. They're doing fine running on commodity servers for their web stuff, and trying to produce their own mobile chips would anger their hardware partners for Android. It might give them a slight edge in the long run, but the short-term harm seems to outweigh that.
Microsoft *might* work. They need some special edge in the tablet war they just jumped into, and AMD is a good match with their successful Xbox line. But AMD isn't known to be particularly good at low-power chips. Perhaps they just haven't tried yet, or some older design could be successfully adapted into tablets (a single/dual-core, low-power K8 paired with a good Radeon design might be a good A6 competitor, especially if Microsoft tries to bill itself both as an 'enterprise' tablet *and* a 'gaming' tablet). But really, although it makes sense for Microsoft to buy some hardware company, AMD isn't the best choice. NVidia might make a better one, but I don't think they're looking to sell out right now.
Samsung might buy parts of the company, but they wouldn't want the whole thing. I imagine they would love the graphics section, maybe some of the CPU engineers, but I doubt they want to enter the full-on CPU market.
You know who might make more sense? Cray, or maybe IBM. AMD stuff is popular for supercomputers, both their Opterons and their FireStream/FirePro cards. IBM isn't too likely (they have enough good hardware people already), but Cray or one of their competitors seems at least more plausible than any of the other suggestions.
Another idea is some gaming company. AMD has a somewhat-competitive graphics division, and a compute side that could handle gaming loads well with some tweaks. Sony is really the most likely - they've *never* been good at the hardware side, only lucking into success with the PS1 and PS2 after some clever business decisions. But I also doubt Sony is smart enough to try to do that, especially since buying AMD might hurt their (Intel-focused) laptop business.
Not quite true. Ever since the Pentium Pro (or somewhere around there), there's been something called "Physical Address Extensions", or PAE. This basically just lets the OS give each process a different 32-bit partition of a 48-bit address space.
32-bit Linux systems generally support it. I believe OS X used it as well on 32-bit processors, before they went purely x86-64.
Windows "supports" it, but doesn't have it enabled by default, and I think they block the option on non-server versions of Windows.
Ironically, I don't think you can. At least, not at all easily.
The FirePro cards shows don't support multiple GPUs driving a single output. So you'd have to find some way to avoid treating the array of screens as a single screen, which is far easier said than done. At the very least, extensive reworking of the game's renderer seems in order, and any game open-source enough for a modder to do that is old or graphically-limited enough that you could drive quad QHD monitors using my laptop.
The Tesla and Xeon Phi don't support video output at all, and the drivers have no OpenGL or Direct3D support to my knowledge. So you would have to reimplement the entire graphics layer, then somehow pipe it out through another video card via shared memory.
And honestly, even then, I'm not sure you'd need a array of eight of these cards for any game on the market. I can max out nearly any game on my laptop, at full 1080p, using the mobile version of the GeForce 640. A mere 384 shader cores. The Tesla K20X is based on the same architecture, but scaled up to 7 times as many cores, and at a clock speed not significantly lower. So a gaming-oriented version of the K20X could max out any game on a three-by-two array of "Full HD" monitors (5760x2160). At four, you're running a three-by-two array of "Ultra HD" monitors (11520x4320) at 120+FPS. There is no way I can think of to feasibly require more from your hardware than this, and that's half the GPUs you proposed.
I realise there's probably an xkcd for this, but floods and "several" seem to be two different extremes to me.
Obligatory XKCD
Not if the rumors of a new Ricochet are true!
Or if it's a completely new franchise, as has been legitimately rumored.
We're not inciting a revolution, we're organizing one. Big difference.
Mainly, we have someone taking minutes.
Did you jump to that conclusion using directions from Apple Maps?