Once again (for those who've either been sleeping or just dumb), Google is a technology company with ad revenue as their major source of income. Not an advertising company. The people who actually create those ads, or the ones you might see on ABC or NBC supporting programming, those are ad companies (well, technically, they're probably media content creators hired by ad companies to create the ads). ABC and NBC are not ad companies, they're media/entertainment companies with ad revenue as their major source of income. As opposed to Home Box Office, Showtime, or Sony Films, which are media/entertainment companies with sales to distributors (Comcast Cable, DirecTV, AMC Movie Theaters) and in some cases, direct to consumer, as their primary sources of revenue. Apple, for example, is a technology company, making a little money in ads, quite a bit in hardware (direct-to-customer retail stores and through distributors) and, of course, as a distributor of media for other companies via the iTunes store. Google is also making money in hardware and direct-to-customer media storefronting for other companies.
The first modern version of Windows, NT 3.0, was release in 1993. The basis of OSX is of course BSD UNIX riding on CMU's Mach kernel. The original Mach project was first working sometime in 1986 or thereabouts, Mach 3.0 was released in 1990. BSD UNIX was first released in 1997. The initial higher level APIs, known as Carbon, evolved out of the original MacOS, which was first released in 1984. Of course, one could go back to the VAX/VMS heritage of Windows NT to 1980, but then again, the BSD UNIX decends from AT&T UNIX, which began in 1969 or thereabouts.
Yeah, Nokia was hurtin'. They had much the same disease that RIM had -- their former smartphone business was being devoured by iPhones and Androids, and rather than even try to compete, they pretty much put head to hole-in-sand and pretended it wasn't happening. So their new phone releases were incrementally better than the year before, which put them a year, then two, behind the hardware that Apple and Google were enabling. Add to that the fact that apps as much as anything drove the consumer smartphone revolution, and both Nokia and RIM sucked at this. Nokia didn't even include the Ovi Store app until much later, and only on some models. Most SymbianOS users though they had feature phones, not smartphones.
But Elop accelerated the fall like made. He made a deal to go Windows Phone exclusive, and to functionally do that a year before he had any Windows Phone hardware to sell. By announcing this, basically announcing that SymbianOS was dead, soon, the whole Nokia income structure came tumbling down. The $1 billion+ losses for each quarter so far this year are due to SymbianOS sales falling off a cliff. Even if Windows Phone had sold well, there's no way that could have made up the difference. Elop didn't HAVE to go exclusive... wouldn't have changed the final results, but it would have given them at least a chance for Microsoft to get their act together. Windows 7 Phone didn't even fix their hardware problem -- Microsoft's use of WinCE in there prevented any competitive hardware from Nokia in 2010-2011. The Lumia 920 is at least in the ballpark, even though this is rapidly turning into the year of quad-core phones...
If not having a good keyboard can be listed as an advantage, I think they have nailed the ASUS Transformer with that one. Odd that ASUS would deliver an actually usable keyboard, rather than plum the depths of history and resurrect the ZX-81 keyboard. Chiclet keyboards, like vertical touchscreens, died a long time ago of natural causes. Microsoft's Jurassic Parking them back into existence won't fix the root failure: people hate these things.
Desktop touchscreen. That's what he's talking about... running the table UI as the primary desktop UI. Microsoft wants you PC to get a touchscreen. Bad idea.
Nice to hear Cook pointing out the fact that vertical touchscreens really don't work. Not just in their testing -- this was a thing, pre-PC, in many of the 70s and 80s CAD workstations. There were touchscreens, light pens, and other "directly interacting with the monitor" input devices. They all failed. It wasn't expense (not in dedicated CAD, prices were so high, paying $1000+ for an interface device would have been lost in the noise), it wasn't functionality (they worked fine)... it was people. We don't like repetitive stress, but particularly on large motor functions. Reaching up, away from your normal comfortable seating position, to touch a large monitor -- just not something that's good for you.
Of course, they wouldn't be Microsoft if they didn't entirely not learn from the past, and actually do it worse. Touch-with-finger screens are inherently a compromise. You wouldn't choose to smear greasy fingers over your viewing device if you could help it.... it's a compromise some are willing to make in order to have an easy to use pocket computer. On the desktop, we use off-screen, horizontally mounted control devices.
But it's clear Microsoft didn't have any cognitive psychologists working on any part of the mess that is The-UI-Formerly-Known-As-Metro, either. This will make one hell of a cautionary tale, though -- hopefully we can stop trying these same kind of stupid ideas on mainstream Linux distros...
Yup... it's easy to see that Apple was inspired by the T|X... pretty much the same screen as the original iPhone. And Palm had mastered the "doesn't slip out of your hand" technology that Apple still has not.
But it wasn't just the hardware or software that Apple did right. It was really the marketing. Palm was fairly convinced that only business people were significant users of PDAs, and only business people would use smartphones -- that's why the Blackberried up the Palm for the Treo devices. Apple was the first major company to go after consumers with a smartphone. And one of the best positioned to break that ice, since they already had that relationship with the consumer via the iPod.
Palm was really kind of clueless in those days. The T|X has the same SOC as the Microsoft Zune, and yet, it didn't even come with a media player (there were a few good ones available). They didn't even bother to apply the SDHC upgrade to the T|X, either... despite supplying that to the Treos that used the same processor. They were really a mess, jerked their customers around, and I was SOOOOO glad to be able to jump off Palm onto a nice, soft landing on Android.
But yeah, what could have been. They even had BeOS, which had already been mutated into BeIA, and would have made a fine basis for a modern smartphone OS.
Also, consider that a large number of recent Windows applications are written in.NET. A.NET program, regardless of the original language used, compiles to an intermediate format, not a machine-specific binary. That last bit, the machine-specific translation, happens at runtime. So if they allowed it, a large number of programs for full Windows could work on Windows RT. But Microsoft doesn't allow it -- only Microsoft can write Win32 applications.
And you're hopefully not all that interested in installing a 16GB applications suite (and that's the compressed size) on a 32GB tablet... particularly given that Microsoft only leaves about 20-something GB free. Windows isn't hidden in NOR Flash as on other tablets, it's a proper part of the flash storage space.
They've had half that problem with Netbooks and Ultrabooks, and people still managed. Though in truth, I've been buying software online for several years... I can't recall the last time I actually bought a CD or DVD... maybe the Windows 7 DVD (I bought four, pre-release)? Plus, Microsoft is pushing and featuring, as the main point of their new Surface TV ad. They snap on this keyboard thing (in a snappy dance number), making the tablet into a laptop, more or less -- that seems to be the marketing hook they're using over the iPad... their $99 keyboard/cover. So "without a keyboard" isn't really an issue. And users know there's a touchscreen, so they don't miss the mouse.
They are absolutely going to expect Windows software to run on a Windows tablet. Unless there's a real informational campaign, done at very consumerly levels, to clarify this.
As I mentioned before, I think Intel's going to be the one to do this, not Microsoft. Microsoft wants to push as many folks onto WinRT as possible, as fast as possible, to get developers interested -- developers know that desktop users are not going to flock to this, and aren't going to accept Metro-only replacements for formerly-usable applications. Forced adoption by iPad-class legions of Windows RT users, however, seals the deal on WinRT.
Intel really, really wants to get x86 into mobile, for the same reasons Microsoft wants to be there. They're fine right now launching x86 into Android for phones, and may go with tablets there too, but it's on Windows tablets they have the advantage: a Z2760 SOC matches ARM SOCs on price, power, and performance. Doesn't beat them, probably doesn't even beat the latest ARM cores on the way, or a quad core A9... but it does run Windows software. About as well as a typical netbook, or worse, but hey, that's better than "won't go".
You do wonder how well these tablets actually will run Windows software anyway. RT stuff is all touch, but not regular Windows. Microsoft is including a pen digitizer (like a Samsung Note or a Wacom) in their Surface Pro for this, but the other guys? That may also be an emerging issue, if you actually need to include a mouse to make regular Windows usable.
The other thing is going to be whether these are seen as real tablets. The ARM and the Z2760 are pretty much in the typical 10" tablet range: under 1.4lbs (the current iPad weighs 1.44lbs), under 0.40" thick (the new iPad is 0.37" thick). Tablets running laptop chips are already being announced, at 1.75-2.0lbs or more and 0.5" thick or more. Then there's battery life... you expect a tablet to go 8-10 hours in typical use. Will fat x86 tablets (many with fans) come anywhere near that expectation? So end users may well find the price and size in RT tablets, but return them after they discover the Windows apps don't work. Then swap them for an Android or iPad -- sure, with exactly the same problem w.r.t. Windows apps, but hundreds of thousands of native apps -- rather than buy the expensive, bulky tablet.
So I bet Intel is pushing very hard on the Z2760, and they'll call it out in ads, in stickers, something in big, bold, candy-like letters letting every consumer know that THIS is the tablet that's still a tablet AND runs real Windows. Only question is whether this all occurred to them, and the stickers are there in the stores on launch day, or whether they'll have the confusion first.
But now that ship has sailed. We have Two versions of "Windows" launching on tablets at the same time, that look the same, but only one runs "Windows Software",
Does a non-technical person buying a tablet even care?
Well, while I can't exactly get inside the head of that non-technical person... why does/would a person buy a Windows computer of any sort? Is it that they're in love with Internet Explorer? That they just can't get enough of Windows Solitare? I'm guessing "no"... it's probably because of software compatibility. I'm a techie... but given my 'druthers, I'd be running HaikuOS or something equally obscure, just to confuse people... if only it weren't for my need to have a bunch of useful applications for actually getting things done. Others pick Windows because of the right games being available only there, etc.
It really comes down to one thing: software. Only software. That's all Microsoft has ever had to offer -- they got an early jump on the competition, did some now-illegal things to quash any possible competition in PC software, and they've pretty much kept that software ball rolling and expanding since 1981. Now they believe they can re-invent 95% of the stuff underneath, still call it Windows, and in a few months, all of that software will show up in WinRT format. Available only in the Windows Store. No actual windows -- it's all full screen. No one will care about being able to use their 30-year investment in software on these new machines, no one will mind having to repurchase all of their software. Seems a bunch of pretty big assumptions there, even for Microsoft.
Yup... and this was a huge benefit to.... not users, but -- oh yeah! Adobe and Avid. Signed FCP users up like crazy last summer, offering 50% competitive upgrades, etc. This was oddly Microsoftian, in that Apple even bent to pressure and put Final Cut Pro 7 back on sale.
That's a good lesson about professional environments -- you really don't mess with them when you can help it. If I buy a new version of Adobe Premiere or Sony Vegas or Altium Designer or AutoCAD, I expect it to be an upgrade for what I'm doing. I'm buying it to get new features of some kind, but I probably don't want a revolution, even when that might be useful two years or more down the road.
Most companies solve this by splitting the line... so Adobe had Premiere, very popular and terribly primitive compared to FCP, Vegas, Media Composer, whatever. They launched Premiere Pro as a new thing... not exactly an upgrade, but a reinvention. The old one was still around for a little while, so users could judge the transition. And they understood their users. Same thing happened when Calkwalk broke 15 years of incremental upgrades to completely reinvent their DAW with Sonar.
Apple, on the other hand, re-wrote from scratch, added some sorely needed features the ailing FCP7 needed (64-bit support, asset managment, etc), but then killed compatibility and eliminated all sorts of professional features. They basically built a great upgrade to iMovie, but a downgrade to FCP7.
Very much what Microsoft is doing in Windows 8. Except for the "sorely needed features" part.
Actually, Microsoft's Surface Pro is based on the i5, not the Atom. That's why the rumors are it'll top a grand in price. And big questions about the weight and battery life.
Intel does have the Atom Z2760 as their go-to chip for tablets. This is x86, 32-bit only, dual core, each core about as powerful as an ARM Cortex A9 at the same speed. But it does have a dual bus memory interface (some ARM SOCs do, some don't), one of the fastest PowerVR GPUs, etc. It'll suck at most traditional Windows apps, but run them, and get priced and stocked right next to Windows RT machines. That's not going to cause confusion?
In fact, I bet it does, but gets solved in the way of competition, and only after a few months of thick consumer confusion. Expect Intel to get out the word, not Microsoft. They'll have some kind of sticker for the x86 based tablets that suggests they're "real" or "full" Windows. Why not -- Intel felt entitled to all of Windows, Microsoft's letting ARM play the game, and so Intel needs some serious payback. Plus, they're at least as hungry as Microsoft for the mobile market -- they apparently see the same writing on the wall, and just as "Windows" is Microsoft's answer to everything (even when they completely change the definition), so x86 is Intel's. Curiously, this Z2760 will be first x86 that's regularly outperformed by ARMs (both quad-core A9s and dual-core Krait, A15, or whatever Apple's using in the A6 SOC) on performance.
And you're not going to get any love from the consumer telling them that, either. They have to update the application they already bought... only, no update's yet available for 3/4 of them. And they have to re-purchase all the rest? That sure starts to make the price of the full Windows 8 tablets (particularly those Atom-based units that'll compete on price with the ARM-based units) even a bargain, if for some reason one must have Windows on a tablet.
Or, looking at it the other way, the only reason you buy Windows is for the applications. Which aren't there yet, on the Windows RT devices. So what's the point? Android and iPads are better supported than RT, unless "Office" is the only application you run. But hey, there are those hundreds and hundreds of wonderful Windows 7 Phone apps that will run on these tablets -- that should make the consumer all bubbles and smiles.
They did this, back in the day, with Win32 on x86, PPC, MIPS, Alpha, etc. Sure, there was no binary compatibility from one CPU to the next. But it pretty much was a simple matter of a recompile, even for device drivers -- the Windows NT HAL actually worked.
And of course, anything written in.NET would presumably "just work" on the ARM systems. That's a pretty big collection of modern Windows programs. But you'd need Win32 to make that possible (well, sure, and.NET libraries and all of that old Windows stuff), and in Windows RT and Windows Phone, only Microsoft gets to use Win32 calls.
For everyone else, Windows RT vs. regular Windows, it's a complete redesign, from Win32 to WinRT. Less work, certainly, if you wrote the app in HTML or something to begin with (not uncommon for Windows UIs), but it's still lots of work. And a complete UI redesign.
Pretty much all of the iPhone apps run on the iPad. Apple's primitive demand for fixed resolutions and bitmaps has certainly fragmented the iOS market -- yeah, you just can't shoehorn an iPad app onto the iPhone. This may well be the same for Windows... Microsoft is making a distinction between Windows, Windows RT, and Windows Phone, and they've said Windows Phone apps (both 7 and 8) will run on Windows and Windows RT, but is it generic WinRT/Metro running on all three, or do you have to code specifically for tablets and phones. At least in Android these days, it's all one thing, and the app adjusts to the device resolution... just as PC apps have been doing for decades.
You would have, if Apple called it a MacOS pad. You also might have been upset if it didn't run iPhone applications, despite being called an iOS device.
Microsoft is saying "Windows" here -- that has a 30 year history of application compatibility, to most consumers. Not to mention that Microsoft (and others) are also building Windows tablets that do run legacy code.
I'm also not suggesting that this is confusing to the typical/. reader... of course it's not. But regular consumers don't even begin to understand these things. I know more than a few people who call their web browser "the internet"... and yeah, I've tried to explain multiple times the difference between a web browser and the internet... doesn't get in.
This is going to be a huge source of confusion. Users are going to be trying very hard to install their existing applications ("how do I get the program from the CD") onto the tablets. They're likely going to be very bent out of shape over having to re-purchase the very few applications that have been ported to WinRT, and worse, angry about those that haven't been. Microsoft has not been clear about this, seemingly intentionally so. Maybe they'll clear things up by the ship date, but I wouldn't hold my breath. Expect record product returns.
What's the point of this. Is Qualcomm the next AMD, in terms of... what? They're certainly not the next x86-drop-in-replacement chip company, if that's what you're looking for. If, instead, you're looking for a new "chimpzilla" to Intel's "chipzilla", Qualcomm has been larger than AMD for years now, particularly made clear now that they're both on equal footing (AMD split off their IC business as Global Foundries, so now, like Qualcomm, they're fabless). And let's not forget Samsung, who's actually beating Intel in IC sales, if not yet market cap.
Samsung clearly has or had an Architectural licence as well... their Hummingbird core was a modified A8, tweaked to run at 1GHz+. This was used in various Samsung SOCs, including the A4 they made for Apple and the chip used in the Galaxy S1. Sure, they used the A9 in recent designs -- so did Apple, until last month. nVidia as well -- they're a known Architecture licensee, and also known to be working on a next generation custom design, but they're using the A9 right now.
Basically, ARM did a much better job on the A9 design, thus, not too many custom versions (the Scorpion from Qualcomm is an A8-class core that runs really fast, not quite an A9).
TI has yet to build their own core -- they've spent most of their efforts on the rest of the SOC. And they're in trouble, at least according to the rumor mill, even looking to sell off their OMAP division.
All of the Qualcomm cores are custom in-house designs. The Scorpion core was fairly similar to the A8, but it ran really fast, so it was sold for these last several years as an A9 alternative. The A9 is about 20% faster than an A8 at the same clock speed, which is why most Qualcomm phones were running at 1.5GHz or so.
The latest Krait design is theirs as well -- that's an "A15 class" core, like the core Apple used in the A6, but not based on the A15. Qualcomm claims about 30% speed improvement over the A9 at the same clock speed, the A15 is expected to run about 35% faster. The Krait is fully ARM compliant, of course, and designed under the Architecture license.
Nope. In fact, it's usually the opposite. Most companies just license cores from ARM. That'll be in the form of a high level, synthesizable processor, GPU, etc. but they're pretty much locked down as-is, few or no modification rights.
Others have the full "Architecture" license, that allows them to make changes to existing ARM cores, or design their own from scratch. Some companies have only done a little with this... Samsung, for example, was one of the companies that made a 1GHz+ ARM Cortex A8 design (the "Hummingbird" core use in the S1 and the Apple A4 chip)... you had to make changes to the original design to run much about 600MHz. Other companies, like Qualcomm, have done much more aggressive in-house designs. They have to be verified as compliant by ARM, it's more work for them, thus this is a higher-priced option.
A bunch of architecture licensees haven't done much obvious with it yet. Samsung did the Hummingbird, but then moved back to bog standard A9 cores for most of their more recent processors. nVidia has an Architecture License, and they're reportedly doing their own ARM V8 processor (64-bit), but their current Tegras are all using standard ARM cores. Microsoft even has an architecture license, but has yet to make any chip -- though given their custom work on the X-Box 360, they certainly have the expertise, even being seen largely as a software company. And of course, Apple not only has an Architecture License, they bought Intrinsity, a company that worked with Samsung on speeding up the A8 (that's kind of their thing, using odd transistor choices, dynamic latches, etc. to speed up chip designs without requiring a die shrink), but it took the iPhone 5 for them to deliver a chip with any custom CPU work.
Slashdot Mirror
Once again (for those who've either been sleeping or just dumb), Google is a technology company with ad revenue as their major source of income. Not an advertising company. The people who actually create those ads, or the ones you might see on ABC or NBC supporting programming, those are ad companies (well, technically, they're probably media content creators hired by ad companies to create the ads). ABC and NBC are not ad companies, they're media/entertainment companies with ad revenue as their major source of income. As opposed to Home Box Office, Showtime, or Sony Films, which are media/entertainment companies with sales to distributors (Comcast Cable, DirecTV, AMC Movie Theaters) and in some cases, direct to consumer, as their primary sources of revenue. Apple, for example, is a technology company, making a little money in ads, quite a bit in hardware (direct-to-customer retail stores and through distributors) and, of course, as a distributor of media for other companies via the iTunes store. Google is also making money in hardware and direct-to-customer media storefronting for other companies.
The first modern version of Windows, NT 3.0, was release in 1993. The basis of OSX is of course BSD UNIX riding on CMU's Mach kernel. The original Mach project was first working sometime in 1986 or thereabouts, Mach 3.0 was released in 1990. BSD UNIX was first released in 1997. The initial higher level APIs, known as Carbon, evolved out of the original MacOS, which was first released in 1984. Of course, one could go back to the VAX/VMS heritage of Windows NT to 1980, but then again, the BSD UNIX decends from AT&T UNIX, which began in 1969 or thereabouts.
In short, they both go back a long way.
Yeah, Nokia was hurtin'. They had much the same disease that RIM had -- their former smartphone business was being devoured by iPhones and Androids, and rather than even try to compete, they pretty much put head to hole-in-sand and pretended it wasn't happening. So their new phone releases were incrementally better than the year before, which put them a year, then two, behind the hardware that Apple and Google were enabling. Add to that the fact that apps as much as anything drove the consumer smartphone revolution, and both Nokia and RIM sucked at this. Nokia didn't even include the Ovi Store app until much later, and only on some models. Most SymbianOS users though they had feature phones, not smartphones.
But Elop accelerated the fall like made. He made a deal to go Windows Phone exclusive, and to functionally do that a year before he had any Windows Phone hardware to sell. By announcing this, basically announcing that SymbianOS was dead, soon, the whole Nokia income structure came tumbling down. The $1 billion+ losses for each quarter so far this year are due to SymbianOS sales falling off a cliff. Even if Windows Phone had sold well, there's no way that could have made up the difference. Elop didn't HAVE to go exclusive... wouldn't have changed the final results, but it would have given them at least a chance for Microsoft to get their act together. Windows 7 Phone didn't even fix their hardware problem -- Microsoft's use of WinCE in there prevented any competitive hardware from Nokia in 2010-2011. The Lumia 920 is at least in the ballpark, even though this is rapidly turning into the year of quad-core phones...
There's always SE Android -- Android for Spooks! No tracking here, thank-you-very-much.
If not having a good keyboard can be listed as an advantage, I think they have nailed the ASUS Transformer with that one. Odd that ASUS would deliver an actually usable keyboard, rather than plum the depths of history and resurrect the ZX-81 keyboard. Chiclet keyboards, like vertical touchscreens, died a long time ago of natural causes. Microsoft's Jurassic Parking them back into existence won't fix the root failure: people hate these things.
Chitty Chitty Bang Bang was great at flying, floating and driving!
And being fictional.
Though the Caractacus Potts character was one of my first inspirations toward engineering, as little kid.
Desktop touchscreen. That's what he's talking about... running the table UI as the primary desktop UI. Microsoft wants you PC to get a touchscreen. Bad idea.
Nice to hear Cook pointing out the fact that vertical touchscreens really don't work. Not just in their testing -- this was a thing, pre-PC, in many of the 70s and 80s CAD workstations. There were touchscreens, light pens, and other "directly interacting with the monitor" input devices. They all failed. It wasn't expense (not in dedicated CAD, prices were so high, paying $1000+ for an interface device would have been lost in the noise), it wasn't functionality (they worked fine)... it was people. We don't like repetitive stress, but particularly on large motor functions. Reaching up, away from your normal comfortable seating position, to touch a large monitor -- just not something that's good for you.
Of course, they wouldn't be Microsoft if they didn't entirely not learn from the past, and actually do it worse. Touch-with-finger screens are inherently a compromise. You wouldn't choose to smear greasy fingers over your viewing device if you could help it.... it's a compromise some are willing to make in order to have an easy to use pocket computer. On the desktop, we use off-screen, horizontally mounted control devices.
But it's clear Microsoft didn't have any cognitive psychologists working on any part of the mess that is The-UI-Formerly-Known-As-Metro, either. This will make one hell of a cautionary tale, though -- hopefully we can stop trying these same kind of stupid ideas on mainstream Linux distros...
Yup... it's easy to see that Apple was inspired by the T|X... pretty much the same screen as the original iPhone. And Palm had mastered the "doesn't slip out of your hand" technology that Apple still has not.
But it wasn't just the hardware or software that Apple did right. It was really the marketing. Palm was fairly convinced that only business people were significant users of PDAs, and only business people would use smartphones -- that's why the Blackberried up the Palm for the Treo devices. Apple was the first major company to go after consumers with a smartphone. And one of the best positioned to break that ice, since they already had that relationship with the consumer via the iPod.
Palm was really kind of clueless in those days. The T|X has the same SOC as the Microsoft Zune, and yet, it didn't even come with a media player (there were a few good ones available). They didn't even bother to apply the SDHC upgrade to the T|X, either... despite supplying that to the Treos that used the same processor. They were really a mess, jerked their customers around, and I was SOOOOO glad to be able to jump off Palm onto a nice, soft landing on Android.
But yeah, what could have been. They even had BeOS, which had already been mutated into BeIA, and would have made a fine basis for a modern smartphone OS.
Also, consider that a large number of recent Windows applications are written in .NET. A .NET program, regardless of the original language used, compiles to an intermediate format, not a machine-specific binary. That last bit, the machine-specific translation, happens at runtime. So if they allowed it, a large number of programs for full Windows could work on Windows RT. But Microsoft doesn't allow it -- only Microsoft can write Win32 applications.
And you're hopefully not all that interested in installing a 16GB applications suite (and that's the compressed size) on a 32GB tablet... particularly given that Microsoft only leaves about 20-something GB free. Windows isn't hidden in NOR Flash as on other tablets, it's a proper part of the flash storage space.
The last version of Cakewalk Sonar I bought was via download... all 16GB worth. So I DLed it at work.
They've had half that problem with Netbooks and Ultrabooks, and people still managed. Though in truth, I've been buying software online for several years... I can't recall the last time I actually bought a CD or DVD... maybe the Windows 7 DVD (I bought four, pre-release)? Plus, Microsoft is pushing and featuring, as the main point of their new Surface TV ad. They snap on this keyboard thing (in a snappy dance number), making the tablet into a laptop, more or less -- that seems to be the marketing hook they're using over the iPad... their $99 keyboard/cover. So "without a keyboard" isn't really an issue. And users know there's a touchscreen, so they don't miss the mouse.
They are absolutely going to expect Windows software to run on a Windows tablet. Unless there's a real informational campaign, done at very consumerly levels, to clarify this.
As I mentioned before, I think Intel's going to be the one to do this, not Microsoft. Microsoft wants to push as many folks onto WinRT as possible, as fast as possible, to get developers interested -- developers know that desktop users are not going to flock to this, and aren't going to accept Metro-only replacements for formerly-usable applications. Forced adoption by iPad-class legions of Windows RT users, however, seals the deal on WinRT.
Intel really, really wants to get x86 into mobile, for the same reasons Microsoft wants to be there. They're fine right now launching x86 into Android for phones, and may go with tablets there too, but it's on Windows tablets they have the advantage: a Z2760 SOC matches ARM SOCs on price, power, and performance. Doesn't beat them, probably doesn't even beat the latest ARM cores on the way, or a quad core A9... but it does run Windows software. About as well as a typical netbook, or worse, but hey, that's better than "won't go".
You do wonder how well these tablets actually will run Windows software anyway. RT stuff is all touch, but not regular Windows. Microsoft is including a pen digitizer (like a Samsung Note or a Wacom) in their Surface Pro for this, but the other guys? That may also be an emerging issue, if you actually need to include a mouse to make regular Windows usable.
The other thing is going to be whether these are seen as real tablets. The ARM and the Z2760 are pretty much in the typical 10" tablet range: under 1.4lbs (the current iPad weighs 1.44lbs), under 0.40" thick (the new iPad is 0.37" thick). Tablets running laptop chips are already being announced, at 1.75-2.0lbs or more and 0.5" thick or more. Then there's battery life... you expect a tablet to go 8-10 hours in typical use. Will fat x86 tablets (many with fans) come anywhere near that expectation? So end users may well find the price and size in RT tablets, but return them after they discover the Windows apps don't work. Then swap them for an Android or iPad -- sure, with exactly the same problem w.r.t. Windows apps, but hundreds of thousands of native apps -- rather than buy the expensive, bulky tablet.
So I bet Intel is pushing very hard on the Z2760, and they'll call it out in ads, in stickers, something in big, bold, candy-like letters letting every consumer know that THIS is the tablet that's still a tablet AND runs real Windows. Only question is whether this all occurred to them, and the stickers are there in the stores on launch day, or whether they'll have the confusion first.
Well, while I can't exactly get inside the head of that non-technical person... why does/would a person buy a Windows computer of any sort? Is it that they're in love with Internet Explorer? That they just can't get enough of Windows Solitare? I'm guessing "no"... it's probably because of software compatibility. I'm a techie... but given my 'druthers, I'd be running HaikuOS or something equally obscure, just to confuse people... if only it weren't for my need to have a bunch of useful applications for actually getting things done. Others pick Windows because of the right games being available only there, etc.
It really comes down to one thing: software. Only software. That's all Microsoft has ever had to offer -- they got an early jump on the competition, did some now-illegal things to quash any possible competition in PC software, and they've pretty much kept that software ball rolling and expanding since 1981. Now they believe they can re-invent 95% of the stuff underneath, still call it Windows, and in a few months, all of that software will show up in WinRT format. Available only in the Windows Store. No actual windows -- it's all full screen. No one will care about being able to use their 30-year investment in software on these new machines, no one will mind having to repurchase all of their software. Seems a bunch of pretty big assumptions there, even for Microsoft.
Yup... and this was a huge benefit to .... not users, but -- oh yeah! Adobe and Avid. Signed FCP users up like crazy last summer, offering 50% competitive upgrades, etc. This was oddly Microsoftian, in that Apple even bent to pressure and put Final Cut Pro 7 back on sale.
That's a good lesson about professional environments -- you really don't mess with them when you can help it. If I buy a new version of Adobe Premiere or Sony Vegas or Altium Designer or AutoCAD, I expect it to be an upgrade for what I'm doing. I'm buying it to get new features of some kind, but I probably don't want a revolution, even when that might be useful two years or more down the road.
Most companies solve this by splitting the line... so Adobe had Premiere, very popular and terribly primitive compared to FCP, Vegas, Media Composer, whatever. They launched Premiere Pro as a new thing... not exactly an upgrade, but a reinvention. The old one was still around for a little while, so users could judge the transition. And they understood their users. Same thing happened when Calkwalk broke 15 years of incremental upgrades to completely reinvent their DAW with Sonar.
Apple, on the other hand, re-wrote from scratch, added some sorely needed features the ailing FCP7 needed (64-bit support, asset managment, etc), but then killed compatibility and eliminated all sorts of professional features. They basically built a great upgrade to iMovie, but a downgrade to FCP7.
Very much what Microsoft is doing in Windows 8. Except for the "sorely needed features" part.
Actually, Microsoft's Surface Pro is based on the i5, not the Atom. That's why the rumors are it'll top a grand in price. And big questions about the weight and battery life.
Intel does have the Atom Z2760 as their go-to chip for tablets. This is x86, 32-bit only, dual core, each core about as powerful as an ARM Cortex A9 at the same speed. But it does have a dual bus memory interface (some ARM SOCs do, some don't), one of the fastest PowerVR GPUs, etc. It'll suck at most traditional Windows apps, but run them, and get priced and stocked right next to Windows RT machines. That's not going to cause confusion?
In fact, I bet it does, but gets solved in the way of competition, and only after a few months of thick consumer confusion. Expect Intel to get out the word, not Microsoft. They'll have some kind of sticker for the x86 based tablets that suggests they're "real" or "full" Windows. Why not -- Intel felt entitled to all of Windows, Microsoft's letting ARM play the game, and so Intel needs some serious payback. Plus, they're at least as hungry as Microsoft for the mobile market -- they apparently see the same writing on the wall, and just as "Windows" is Microsoft's answer to everything (even when they completely change the definition), so x86 is Intel's. Curiously, this Z2760 will be first x86 that's regularly outperformed by ARMs (both quad-core A9s and dual-core Krait, A15, or whatever Apple's using in the A6 SOC) on performance.
Should be an interesting few months...
And you're not going to get any love from the consumer telling them that, either. They have to update the application they already bought... only, no update's yet available for 3/4 of them. And they have to re-purchase all the rest? That sure starts to make the price of the full Windows 8 tablets (particularly those Atom-based units that'll compete on price with the ARM-based units) even a bargain, if for some reason one must have Windows on a tablet.
Or, looking at it the other way, the only reason you buy Windows is for the applications. Which aren't there yet, on the Windows RT devices. So what's the point? Android and iPads are better supported than RT, unless "Office" is the only application you run. But hey, there are those hundreds and hundreds of wonderful Windows 7 Phone apps that will run on these tablets -- that should make the consumer all bubbles and smiles.
But it's more than that.
They did this, back in the day, with Win32 on x86, PPC, MIPS, Alpha, etc. Sure, there was no binary compatibility from one CPU to the next. But it pretty much was a simple matter of a recompile, even for device drivers -- the Windows NT HAL actually worked.
And of course, anything written in .NET would presumably "just work" on the ARM systems. That's a pretty big collection of modern Windows programs. But you'd need Win32 to make that possible (well, sure, and .NET libraries and all of that old Windows stuff), and in Windows RT and Windows Phone, only Microsoft gets to use Win32 calls.
For everyone else, Windows RT vs. regular Windows, it's a complete redesign, from Win32 to WinRT. Less work, certainly, if you wrote the app in HTML or something to begin with (not uncommon for Windows UIs), but it's still lots of work. And a complete UI redesign.
Pretty much all of the iPhone apps run on the iPad. Apple's primitive demand for fixed resolutions and bitmaps has certainly fragmented the iOS market -- yeah, you just can't shoehorn an iPad app onto the iPhone. This may well be the same for Windows... Microsoft is making a distinction between Windows, Windows RT, and Windows Phone, and they've said Windows Phone apps (both 7 and 8) will run on Windows and Windows RT, but is it generic WinRT/Metro running on all three, or do you have to code specifically for tablets and phones. At least in Android these days, it's all one thing, and the app adjusts to the device resolution... just as PC apps have been doing for decades.
Only after you crack the bootloader. At least they're only using SH-1 for authentication....
You would have, if Apple called it a MacOS pad. You also might have been upset if it didn't run iPhone applications, despite being called an iOS device.
Microsoft is saying "Windows" here -- that has a 30 year history of application compatibility, to most consumers. Not to mention that Microsoft (and others) are also building Windows tablets that do run legacy code.
I'm also not suggesting that this is confusing to the typical /. reader... of course it's not. But regular consumers don't even begin to understand these things. I know more than a few people who call their web browser "the internet"... and yeah, I've tried to explain multiple times the difference between a web browser and the internet... doesn't get in.
This is going to be a huge source of confusion. Users are going to be trying very hard to install their existing applications ("how do I get the program from the CD") onto the tablets. They're likely going to be very bent out of shape over having to re-purchase the very few applications that have been ported to WinRT, and worse, angry about those that haven't been. Microsoft has not been clear about this, seemingly intentionally so. Maybe they'll clear things up by the ship date, but I wouldn't hold my breath. Expect record product returns.
What's the point of this. Is Qualcomm the next AMD, in terms of ... what? They're certainly not the next x86-drop-in-replacement chip company, if that's what you're looking for. If, instead, you're looking for a new "chimpzilla" to Intel's "chipzilla", Qualcomm has been larger than AMD for years now, particularly made clear now that they're both on equal footing (AMD split off their IC business as Global Foundries, so now, like Qualcomm, they're fabless). And let's not forget Samsung, who's actually beating Intel in IC sales, if not yet market cap.
Samsung clearly has or had an Architectural licence as well... their Hummingbird core was a modified A8, tweaked to run at 1GHz+. This was used in various Samsung SOCs, including the A4 they made for Apple and the chip used in the Galaxy S1. Sure, they used the A9 in recent designs -- so did Apple, until last month. nVidia as well -- they're a known Architecture licensee, and also known to be working on a next generation custom design, but they're using the A9 right now.
Basically, ARM did a much better job on the A9 design, thus, not too many custom versions (the Scorpion from Qualcomm is an A8-class core that runs really fast, not quite an A9).
TI has yet to build their own core -- they've spent most of their efforts on the rest of the SOC. And they're in trouble, at least according to the rumor mill, even looking to sell off their OMAP division.
All of the Qualcomm cores are custom in-house designs. The Scorpion core was fairly similar to the A8, but it ran really fast, so it was sold for these last several years as an A9 alternative. The A9 is about 20% faster than an A8 at the same clock speed, which is why most Qualcomm phones were running at 1.5GHz or so.
The latest Krait design is theirs as well -- that's an "A15 class" core, like the core Apple used in the A6, but not based on the A15. Qualcomm claims about 30% speed improvement over the A9 at the same clock speed, the A15 is expected to run about 35% faster. The Krait is fully ARM compliant, of course, and designed under the Architecture license.
Nope. In fact, it's usually the opposite. Most companies just license cores from ARM. That'll be in the form of a high level, synthesizable processor, GPU, etc. but they're pretty much locked down as-is, few or no modification rights.
Others have the full "Architecture" license, that allows them to make changes to existing ARM cores, or design their own from scratch. Some companies have only done a little with this... Samsung, for example, was one of the companies that made a 1GHz+ ARM Cortex A8 design (the "Hummingbird" core use in the S1 and the Apple A4 chip)... you had to make changes to the original design to run much about 600MHz. Other companies, like Qualcomm, have done much more aggressive in-house designs. They have to be verified as compliant by ARM, it's more work for them, thus this is a higher-priced option.
A bunch of architecture licensees haven't done much obvious with it yet. Samsung did the Hummingbird, but then moved back to bog standard A9 cores for most of their more recent processors. nVidia has an Architecture License, and they're reportedly doing their own ARM V8 processor (64-bit), but their current Tegras are all using standard ARM cores. Microsoft even has an architecture license, but has yet to make any chip -- though given their custom work on the X-Box 360, they certainly have the expertise, even being seen largely as a software company. And of course, Apple not only has an Architecture License, they bought Intrinsity, a company that worked with Samsung on speeding up the A8 (that's kind of their thing, using odd transistor choices, dynamic latches, etc. to speed up chip designs without requiring a die shrink), but it took the iPhone 5 for them to deliver a chip with any custom CPU work.