Hasn't this been known for about 6 months now? Clover Trail isn't an Intel part, it's a PowerVR core licensed from Imagination. Intel doesn't have the rights to release the specs on PowerVR and Imagination has never wanted an open source implementation, even if Intel writes it.
1) Microsoft has exerted leverage over Intel and got them to go down the preferred supplier route. Possible, and not overly surprising.
2) Intel is trying to carve a niche out and thinks its best bet is with the old hound.
3) The boring technical reason that there's loads of implementation and they can't be bothered making the changes to the Linux packages.
It sounds like it'll run but with none of the fancy features to shut down cores under low use, but Linux developers have been adding missing functionality for years so it's not exactly unusual.
Option 3 has no basis since Intel doesn't need to write a single line of code to get their new chip supported under Linux, they just need to release the specs. Which brings us to:
4) Just like many graphics chips, this processor does some black magic to boost performance and Intel doesn't want to give any clues to competitors by releasing specs and/or open drivers.
Regarding "Just like many graphics chips, this processor does some black magic to boost performance and Intel doesn't want to give any clues to competitors..."
One reason it's common for GPU makers to be secretive is that there is a very big risk that GPUs and their drivers violate patents, and the makers of these products want to make it as difficult as possible to ascertain whether patents are being violated. They won't even broach the topic because that would draw them in to willful infringement.
So a proprietary OS maker like Microsoft can conclude a deal with IMGTek and give them all the secrecy they want, and Linux is left out.
Apparently even Intel can't get IMGTek to open up. Intel probably doesn't want to complicate their negotiating position with IMGTek by saying anything about whether Intel's GPUs might be sufficiently power-efficient to go into mobile parts. There is a lot of money, and a lot of product viability at stake. Intel GPUs have open source drivers.
> One reason it's common for GPU makers to be secretive is that there is a very big risk that GPUs and their drivers violate patents, and the makers of these products want to make it as difficult as possible to ascertain whether patents are being violated. They won't even broach the topic because that would draw them in to willful infringement.
This is another example of how patents are directly counterproductive to their originally intended goal of encouraging open innovation. Patents are not just creating a minefield, but are encouraging everyone to be more secretive at the same time.
Most likely not performance as in rendering performance, but as in power saving; wasn't this also NVIDIA's reason for not releasing specs to Linux developers?
I actually have some sympathy with a hardware company in this position; if all your competitive advantage is in actually in software, open source is a tricky proposition. Nobody cares about the driver interface to the rendering pipeline any more - I imagine it's heavily dictated by the need to make open standards fast (eg. OpenGL). Power saving? That's a whole different ballgame.
It's a real pain that there are no useful standards for power saving - I've had to disable almost all of the cool processor features on my DAW (C-states, core parking, EIST) because they completely wreck real-time audio latencies, and cause clicks and glitches in DAW software (using external FireWire audio hardware). I notice that Cubase on Windows 7 dynamically installs a different power saving profile on startup, but presumably that means running it as Administrator, and we're back to the bad old days of running as root everywhere.
Intel had the power to ask for that in the negotiations and did not. Deciding to still ship the product in these conditions is therefore purely an Intel decision.
The future SoC that's expected to introduce the new graphics is "Silvermont" and expected availability by early 2013. Before the exciting Silvermont there is expected to be the "Clover Trail" Atoms released this calendar year, but it doesn't sound like that SoC will bear new graphics capabilities. I hope to have some more details soon and am very excited to see Intel do away with the PowerVR graphics and its horrific driver support.
Android has made Nokia nearly bankrupt already, Siemens/BenQ has stopped making mobile phones while Android existed. Windows CE/Mobile/whatever seems to have no chance on the market anymore, eventhough they seem technical superior! Palm completely failed with its WebOS.
In fact Blackberry is the last survivor of the old cellphone world. (EDIT: And Samsung of course, but they are just copying Apple stuff and get sued for it.)
The only ones having profit from Android are Google and HTC. Google because they created the system and run the App Store. HTC because they manufacture the reference devices and because they are cheaper than the rest.
Wake up mates, open source destroys the old business models.
Google made only $550 million from Android between 2008 and 2011 according to figures from the Oracle trial (http://www.guardian.co.uk/technology/2012/mar/29/google-earn...), compared to $38 billion in total revenues for 2011. While half a billion over three years is not trivial, it is rather insignificant compared to $38 billion over one year for search advertising. Trefis analysis of Google's stock price estimates the value of non-Motorola Google Phone properties at 2.38% of their total stock value (http://www.trefis.com/company?hm=GOOG.trefis#/GOOG/n-1336?fr...). To use Warren Buffet terms (as per an analysis by Bill Gurley via Erick Schonfeld), Android is not Google's castle, it's their moat. (http://techcrunch.com/2011/03/25/search-googles-castle-moat). The point of Android (and to a lesser extent, Chrome OS) isn't to generate revenue itself but to protect search revenue (if, for example, iOS were to switch to Bing by default). Given that Apple CFO Peter Oppenheimer had stated in 2010, that iOS's App Store ran at about the breakeven point, and was therefore not a significant source of revenue for Apple, (http://appleinsider.com/articles/10/01/26/app_store_wildly_s...) it is not wholly unbelievable that Google Play would also be an insignificant source of revenue for Google (especially given the ratio of free apps to paid apps is higher on Google Play than on the App Store).
As for HTC: "HTC Corp. said Friday its unaudited second-quarter net profit fell 58% from a year earlier, as the Taiwanese smartphone maker struggles to compete with industry leaders Apple Inc. and Samsung Electronics Co. HTC said in a statement its unaudited net profit for the three months ended June 30 was $7.40 billion New Taiwan dollars ($247 million), down from NT$17.52 billion a year earlier. Its revenue dropped 27% to NT$91.0 billion from NT$124.40 billion." (http://online.wsj.com/article/SB1000142405270230414120457751...).
You are also incorrect about HTC manufacturing the reference devices. While they did make the Google Nexus One, the Google Nexus S was Samsung, the Google Galaxy Nexus was Samsung, the Motorola Xoom was technically the Honeycomb reference device (read as: Nexus), and the Google Nexus 7 is Asus. In other words, HTC has manufactured only one of the five Nexus devices (out of six if you count the Nexus Q). To assert that HTC is generating profit "because they manufacture the reference devices" is inaccurate given that the Nexus One was released in 2010 and no longer sold through first party sources such as Google Play (the devices available for purchase at the time of this writing are the Galaxy Nexus, the Nexus 7, and the Nexus Q).
There is, however, someone making a profit off of Android: Samsung. From the same article: "In stark contrast to HTC's weak earnings, Samsung said earlier Friday that it will likely post a record quarterly operating profit for the second quarter that ended June 30. The South Korean electronics giant, which is due to release audited results later this month, expects an operating profit of between 6.5 trillion won ($5.7 billion) and 6.9 trillion won for the quarter, compared with 3.75 trillion won a year earlier." (http://online.wsj.com/article/SB1000142405270230414120457751...)
TLDR: The assertion that Google and HTC are (significantly) profiting from Android is incorrect and instead should assert that Samsung is the only one making (significant) profits from Android.
However those numbers are not everything. What counts is whether customers buy products or not. There are plenty of examples of successful mass products that were sold too cheap. For instance Microsoft's X-Box, they sold it much under price. You might list that under marketing costs. Their strategy however paied off, now they are a big player in the game device market.
Of course Google spend lots of money into Android, we don't want to imagine what all this patent crap and marketing costs. Android is however the predominant mobile platform and HTC its biggest vendor. (With reference devices I mean the actual devices that were supposed to be used by developers.)
Samsung is without question in a nice position but they are replacable, they are a big brand and they have a lot to loose. HTC appeared out of nowhere and is now one of the biggest players in the smartphone market. Just ask yourself who the real profit makers are.
Samsung is doing short-term profit. You know why? Because people don't by Samsung devices, they buy Android devices. It's not Apple vs Samsung vs HTC, it's iOS vs Android.
Going back to the actual topic: who cares what processor is inside an Android? Only the CPU manufacturers. The customers and the assemblers just care about software support and performance.
HTC is not Android's "biggest vendor." Samsung Telecommunications has more device sales, more revenues, and more profit than HTC.
Nexus devices are the reference devices that were supposed to be used by developers: at the Google I/O conference, each attendee received a Nexus 7, Nexus Q, and Galaxy Nexus (http://www.computerworld.com/s/article/9228585/Google_gives_...). If you have a different opinion of what Android "reference devices" are, then provide them for evaluations by the denizens of Hacker News (of which includes a number of Google employees).
It is not necessary to "ask yourself who the real profit makers are," because the revenue numbers for HTC and Samsung are available to the public. The numbers state that Samsung is alreaduy generating more profit than HTC by a significant margin. Samsung's numbers have been increasing. HTC's numbers have been decreasing. Here are some articles that perfectly summarize the point:
"Soaring sales of smartphones lifted Samsung Electronics Co.'s profit to a company record in the first quarter and executives said that a new model hitting dealers next month will fuel its financial results during the second." (http://online.wsj.com/article/SB1000142405270230472330457736...)
Yeah yeah yeah, I might have been wrong with some details and with the sales numbers in 2012. However my main point stays the same: the traditional cell phone manufactures are in big trouble because of Android.
Samsung may be exceptional, although I doubt their current success is long-term. HTC's success, eventhough it is in absolute numbers smaller than Samsung's, is way larger in relative numbers. Samsung was popular before Android, HTC was when Android was released just a "chinese noname brand".
To illustrate this even more: HTC was founded in 1997, Samsung was founded in 1938.
There a many examples of companies that are able to make lots of money with hypes, but most often after the hype is over, they stop earning that extra money. Even when the current smartphone hype is over, HTC earned something that Samsung already had: brand recognition.
Nokia also made a similar amount in patent licensing from Apple. I believe Nokia is also suing HTC so might start collecting patent licensing from them at some point.
I'd bet on 3, too. Sounds like they slapped the only good-enough gfx core they are sure will work on an Atom, so they'd have something that at least works well with Windows 8 at launch. The next-gen Atom cores will use the same gfx core as their Ivy Bridge processors, but that won't be ready for some time.
Much like the Medfield SoC it derives from, I suspect Clover Trail doesn't use a BIOS or various basic I/O lines and conventions that have been standard in the PC world for decades. So yeah, I wouldn't expect a stock Win7 to run at all.
It's not that the chip won't run linux. It's that the device it comes shipped in will be UEFI locked down such as to make it impossible to install anything...
Until it's jailbroken and you install whatever on it you like.
This will be interesting to watch. There is a whole generation of users that do not use Windows as their primary OS. The brand power that it once had is disappearing. Users can now pick between Linux and iOS for most of their computing needs. The failure of Windows Phone to compete on that market shows that there is a significant amount of mindshare being lost. My guess is that MS is trying to lock people into using the OS again, but this time it won't happen. There are too many options right now. Users are so much more educated than before. Every kid out there can root their Android and upload a custom ROM. When they see their smartphone (or pocket PC) being locked down, they will just get something that doesnt lock them. All of that and Apples marketing power. MS lacks it.
A better way of putting it is that there is a generation of users that do not differentiate between hardware and software. They just know devices.
And if you are only supplying one half of the device, you are in most respects going have trouble sustaining your business in 5-10 years time because the market for just the OS or just the hardware will be tiny or have tiny margins.
Intel's decision to support only Windows 8 on Clover Trail might work for laptops but seems very risky for tablets, where x86 tablets running Windows 8 look to be priced close to Apple's Ipad and significantly higher than Google's Nexus 7 and Amazon's Kindle Fire HD.
Uhm, they surely meant the other way around?
Practically no one have ever switched OS on a tablet/phone whereas on a laptop you'd certainly expect to be able to run linux.
And what's the point of talking about the price? An x86 windows tablet is the only tablet that is able to run real applications. For those that want that the price is more than justified. Those that don't want that should probably buy an arm + windows RT tablet (which is probably more competitive to the reduced feature set of smartphone-based devices such as android and iOS).
"on a laptop you'd certainly expect to be able to run linux."
The fraction of end users to whom it would occur to consider Linux on their laptop is tiny. The fraction of consumers who would consider it is even smaller. The fraction of consumers who would expect their new laptop to support Linux is negligible. The expectations of the HN community are atypicle.
And regardless, you can, rightly, expect a laptop to be able to run linux (albeit you can't expect it to be quirk-free).
You can run linux and other operating systems on many phones and tablets as well. But the fraction of those that consider that is waay smaller than for a laptop (and I doubt anyone would expect it from a typical phone/tablet). As well as the result being barely usable and more of a proof of concept in the very most of the cases.
To clarify: Based on current trends regarding hardware, e.g. the "post PC era", and software, e.g. "walled gardens"; an expectation that in the future one will be able to run the OS of their choice on common computing devices with the sanction of the hardware manufacturer increasingly looks less reasonable.
Yes, it is possible to boot Linux on an iPad. And it is also likely that one will be able to likewise boot Linux on devices using the new Intel parts.
But this does not make for a reasonable expectation that end users (let alone general consumers) will have a viable practical alternative to do so in the future. In general the trends are in the other direction, particularly in regards to hardware being configured as general computing devices rather than having systems more firmly embedded.
A few years ago, Napster, made it look as if music would be freely distributable. A few years before that, there was no advertising on the internet. Earlier, commercial software was shareable among people who did not purchase it. At one time computers filled large rooms.
The trick is that there are two different sorts of "expect".
Suppose a parent finds their teenage kid watching TV instead of studying for an exam and says, "I expect you to do well on that exam tomorrow!". Do they expect their child to do well? Yes. Do they expect their child to do well? No, they would be foolish to.
At the present, both types of "expect" apply to this discussion. Some companies seem intent on eliminating one of those expectations, but they cannot do anything about the other.
In other words, the same "Windows 8 Mobile will come with a locked EFI Secure Boot" we've known about. The OS is signed and verified by the bootloader.
When I read this, I realized that the author was supporting a preconceived conclusion. Apple uses ARM, in no small part because they bought a fan not long ago; and, of course, iOS is not out in the wild to be installed on hardware from OEM's such as Acer or HP.
"The company was founded as Advanced RISC Machines, ARM, a joint venture between Acorn Computers, Apple Computer (now Apple Inc.) and VLSI Technology."
Where would Intel be without Microsoft today??? They need windows to continue to succeed. ARM chips are the enemy of Intel, thus IOS and Android/linux are the enemy of Intel.
This will never fly... That article you linked says that the Intel based phones will only run 90% of android apps. Who would choose one of these phones over an arm based one? It would take a lot of incentive for someone choosing between them. The cost would have to be significantly cheaper or the phone significantly better. I see a non-starter here.
