Google needs to announce the ARM Powered Google TV to compete with the rumored Apple iTV. Google TV could be sold for less than $99 using ARM instead of $299 using Intel.
Here's a recent demonstration of Google TV showing its integration with Dish Networks services as reported by Engadget.com:
What hardware requirements will the ARM Powered Google TV have? ARM cortex A9 with DDR3 at the minimum for fast HD 1080p resolution browsing and interfaces? HDMI input and output (pass-through)? IR-blaster?
My guess is that there could be 2 different versions of Google TV for ARM Powered devices:
1. The full version: with full integration with existing cable/satellite box (HDMI in/out and IR blaster) plus same functions as the basic version.
2. The basic version: that only does the IPTV, Internet-only and media streaming features.
Using ARM, the Internet-only basic Google TV version could be sold at or below $49 while the fully backwards compatible with cable/satellite channels Google TV experience could be sold at or below $99.
I don't think Google has signed any "exclusive" partnerships with Intel, they have a partnership for sure, just as Google has a partnership with Intel powering the more than a million Google servers that are out there. Intel feels left out of the whole Android ecosystem, so they are the ones who have been most desperate to at least be a part of the Google TV initiative. I think it's more about Google waiting for the next generation ARM Cortex A9 to be ready to support full HD 1080p interfaces smoothly before they announce ARM support. Early next year, Google TV will be open sourced anyways, so by that time all the ARM vendors will have it.
Google owns the most servers of any one company out there. Literally, it is rumored to be in the millions. Youtube streams out more than 2 billion video views per day. If an average video view is 30 megabytes, that means Youtube uses at least 60'000 Terrabytes of bandwidth per day. As I am not sure where to find the current estimate of global bandwidth capacity to all home Internet lines, let us compare that with the current reported trans-atlantic bandwidth capacity at around 10 to 20 Terrabit/s, that means a maximum bandwidth between Europe and the USA of 108'000 Terrabytes per day. 2009 price per trans-atlantic 10Gbit/s was $14'000 per month. If Google had to transfer half of Youtube's daily bandwidth over the atlantic, that would cost $11 Million per day in bandwidth costs. Obviously, that scenario would be totally impossible.
The only way Youtube can provide up to 1080p and even 4K video to everyone, for free, on demand, is by cleverly distributing all the most popular videos on many, hundreds, maybe thousands of sort of caching servers that deliver the Youtube videos to every ADSL, Cable, Fiber to the home users around the world. What Youtube does, is that it tries to deliver the video from a Youtube cache as close to the user as possible. Most likely, that for the past years already, Google has been negociating with all the worlds major ISPs to build those Youtube caching servers right onto the local fiber optic backbones and server centrals of every major ISP.
This means, Google already has had to pay ISPs and major backbone providers for preferential access to its contents. This is the only way Google can provide a service that is fast enough for everyone to experience full bandwidth and fast buffering on all these high quality videos and have it all work smoothly in prime time, and when everyone wants to watch the same popular videos.
I'd like bloggers who critisize Google for it's stance on Net Neutrality to consider that Youtube may account for about half of current global web bandwidth consumption excluding peer-to-peer like BitTorrent. And that Youtube's share of global bandwidth consumption can only increase fast as Google TV comes out. What I think is the core issue we have to consider, is that it would be the worst if ISPs somehow because they might be more interested in staying with the status quo, that they may prevent an expansion of web video distribution by limiting the way high quality videos can be distributed freely on-demand to everyone.
While I think it is important to make sure startups don't have any barriers to present their new ideas and new technical solutions to the web, I also think it is kind of logical that a startup won't be able to afford distributing HD quality video as smoothly as Youtube does it, a startup cannot afford to buy thousands or even millions of servers and install them on every ISP and backbone around the world. AS once the cache server is embedded directly onto the ISPs backbone, that bandwidth, no matter how much it is, costs very little for the ISP to provide, at least compared to a scenario where the ISP would have to fetch all that data from the other side of an ocean.
I want to see a scalable solution that basically allows Youtube to expand even much further and provide even higher quality. It would be great if 720p 2mbit/s HD, 1080p 4mbit/s HD and even 24mbit/s 4K was streamed at full speed to every user. And imagining a scenario where a larger and larger share of everyone's 5 hour daily TV watching did stream on-demand from the web. This means that the current 60 Petabytes/day that Youtube consumes today could turn into 6000 Petabytes/day within months.
As for Net Neutrality on wireless networks. I have often tried to ask Telecom companies how much bandwidth there really is in those base stations. The answer I kind of got, was that the bandwidth is not much, maybe only in the tens of megabit/s per base station. So obviously, it may not really matter what generation is used, be it 3G, 3.5G, 3.5G+, 4G/LTC, WiMax, as soon as more than 2 or 3 people start to want to have several megabit/second in an area of a neighborhood or so, the technology just won't be able to handle it all. So for sure, I think Wireless networks need to be regulated. Especially as there isn't really enough bandwidth for everyone to stream video on 3G nor LTE networks. Most importantly, I want to see regulation enforce that bandwidth for VOIP usage may not be blocked or down-graded compared to the exact same quality of the bandwidth used for "normal" voice calls.
The only solution to the bandwidth limitations of wireless networks, is to install smaller base stations closer to every user, so called micro-cells, also called Fem2cells. I believe the best spectrum that we should all install those base stations for would be the 700mhz white spaces spectrum, that is why I would like to hear someone say that it would be possible to build a FON.com like network using unlicensed and free to use White Spaces 700mhz. Everyone would install a $20 White Spaces micro-cell routers onto their home wireline ADSL/Cable/Fiber connections, sharing the bandwidth into their neighborhoods, using one same browser based access authentication system worldwide, that will provide enough wireless bandwidth for everyone to do whatever they want.
Americans watch in average 5 hours of TV every day. Imagine a revolutionary $99 set-top-box which you add to your living room. This one increasingly brings more and more content from the Internet to the HDTV. This one even improves the experience of regular TV channels by overlaying search features and better targeted ads (which can finance better TV content).
Engadget is reporting on the rumored $99 Apple iTV set-top-box. It will basically be like an iPod Touch, without the screen and with an HDMI output and a remote control. It'll have the latest Apple A4 processor which is based on the Hummingbird 45nm ARM Cortex A8 processor (similar to the one used by Samsung in the Galaxy S) designed by Intrinsity before they were bought by Apple. As usual, I don't expect Apple to include support for many video and audio codecs and a proprietary iTunes synchronization over the network is more likely than support for the Samba and Upnp local file sharing standards.
The idea here is that by using the optimal ARM processor of the market, a very powerful yet very cheap set-top-box can be made. One that brings full 720p web browsing to the HDTV, but also re-designed and optimized graphics accelerated user interfaces to the HDTV, basically smooth interfaces for Youtube and other video-on-demand sources, to thus be watched directly on the HDTV.
I've video-blogged about Android based set-top-boxes such as the $50 design by Webia Technologies and Bonux and the $129 (retail target price) one made by Keenhigh mediatech. Both can run the latest Android 2.2 software (when available) with full 3D graphics acceleration even though their processors are likely ARM9 or ARM11 based.
As Google goes along partnering with Intel to release Google TV soon, I expect the Intel based designs to be sold at $199 or likely above that. I think it would be nice to know how soon the customized Android software that represents the Google TV disribution would also be optimized for use on cheaper ARM Powered solutions. As Android on those cheap prototypes looks great, it would be good for those devices to know they can rely on a Google OS optimized for use with a remote control and optimized for easy access to revolutionary HDTV features. Including the support of Youtube in HD quality on all those cheap boxes.
The basic hardware features needed for full Google TV support on cheap ARM Powered set-top-boxes I think are HDMI input and output (pass-through) for overlaying features to contents from existing Cable/satellite set-top-boxes as well as the IR blaster to control that other set-top-box. But for Internet and media streamer features only, all that is needed is just a Google TV for ARM software release.
According to market analytics company Canalys second quarter 2010 smart phone sales report, Android is the fastest selling smart phone platform in the USA between April and June 2010, with 34% of the sales in the US market, in front of Blackberry RIM who sold 32% of smart phones and Apple iOS who sold 21%.
Canalys’ detailed, globally consistent data shows it is the collective growth of Android device shipments across a range of handset vendors’ portfolios that is most remarkable. With key products from HTC, Motorola, Samsung, Sony Ericsson and LG, among others, shipments of smart phones running the Google-backed Android operating system grew an impressive 886% in Q2 2010.
As I posted in my previous post Why are we still waiting for the sub-$250 Android super phones?, this market domination in the USA is reached by the combined Android vendors without them even starting to provide much cheaper Android phones. As in my opinion, the strength of Android is not only the differentiation and increased number of features and choices, I think it is also most importantly the opportunity for competition in the smart phone industry to bring lower unlocked smart phone prices. Once Android super phones are sold below $250 unlocked, and that those can be used for pre-paid plans without long term contracts, I believe that the Android market share will even further increase.
New Android super phones are arriving on the market every few weeks, promoted by all carriers and selling twice as fast as the iPhone. We are also seeing a few lower cost Android phones being sold with smaller resistive screens, using slower processors. Specialist sites like isuppli.com have analyzed the Bill Of Material of all those phones and they are saying that even the top of the line of Android super phones can be manufactured for $165 or below. My question is then still, as we are seeing more and more competition in the Android market, when are we going to be able to buy all those Android super phones at below $250, with unlocked 3G and 4G SIM card slots and no obligation to sign up for 2 years of very expensive, often over $2000 of subscriber contracts?
Before the official release of the Nexus One, I had speculated that it may be sold by Google directly to consumers below $200, as I thought that Google wouldn't be interested in making a profit on the hardware as their strategy is to make a profit on advertising and online services. With probable pressure from manufacturers such as HTC, and from pressure by the big carriers, Google has not been allowed to yet introduce such a disruptive business model to the Android super phone market, at least not yet. It's the same thing Andy Rubin told me off camera at Mobile World Congress, that Google wasn't the one deciding what should be the pricing of the Nexus One.
It's not only consumers in Europe and the USA who I think would be glad to get cheaper unlocked Android super phones, I believe there is a gigantic market available right there to the Android super phone manufacturers if they would aim to provide sub-$250 unlocked Android super phones to the people of developing countries. Each year, over 1 billion phones are bought by consumers worldwide, in fact, mostly in developing countries right now. Those handsets are mostly low cost Nokia phones and the likes. Super phones may be reaching only 100 million units sold each year so far. But that could easily double or triple from one year to the next, as soon as manufacturers decide to sell them at more reasonable cost to the consumers who want unlocked or who only care for pre-paid mobile plans. Most of the world's nearly 5 billion mobile phone users are using pre-paid plans.
Shanzai.com recently published an editorial analyzing the upcoming wave of sub-$100 Android phones that could be coming based on the new low cost ST Ericsson 416MHz T6719 platform that I filmed in the Acer beTouch E120 and Acer beTouch E130 at Computex. Marvell may also be preparing a lower cost Android super phone platform as demonstrated in the O Phones line of Android super phones that is sold only in China for now. Texas Instruments, Qualcomm, Freescale and others surely have some low cost Android super phone platform plans that they must be preparing.
The issue is the conflict of interest that there might be between Android super phone's current customers, which are the major European and American carriers, and the interest of consumers worldwide to see lower cost Android super phones with less constraints, less obligations and less contracts. Many consumers in Europe and the USA are paying the equivalent of 3 or 4 months rent every year just for their mobile phone contracts, which is insanely expensive when you think about it.
Wouldn't it be nice if one could get a $250 unlocked Android super phone, with all the latest capacitive screens and fastest available ARM processors, also with 2 SIM card slots, one for eventual pre-paid or subscription based voice/sms SIM card, and the other for an eventual data-centric 3G or 4G service. On top of which it would still have all the WiFi and Bluetooth features, as well as video-chat that works on any network, and fully unlimited and unlocked features of tethering, Mifi function and HD video streaming, also coming on any networks. All that with HDMI output if needed and HD camcorders right there inside. Then also, sub-$125, there would be smaller resistive screens and lower performance ARM and single SIM normal devices but also unlocked and usable for pre-paid plans. Shanzai.com and Visionmobile.com speculate that advanced features can be included on the cheaper Android hardware using Mobile Virtualization.
Virtualization is new to mobile, but established in the data center, fundamental in cloud computing and increasingly popular on the desktop.
Mobile Virtualization lets handset OEMs, operators/carriers and end-users get more out of mobile hardware. It decouples mobile OSes and applications from the hardware they run on, enabling secure applications and services on less expensive devices today and deployment on advanced hardware tomorrow.
Without much of the cheap Android phones yet on the market, already Android is selling twice as fast as the iPhone. Imagine how much faster they will sell once the cheaper Android super phone prices start to be released. Perhaps a bit more competition in this market is needed to trigger this, perhaps manufacturers and carriers will be careful not to cannibalize their huge profit margins in Europe and USA while some of them will launch hundreds of millions of cheaper Android handset options in the developing countries.
Engadget and a bunch of other blogs have been reporting these last few days about the cool Augen branded Android Smartbook and Tablets that are being released in the US market at affordable $99 and $149 prices by Super Market chain KMart. I just would like to remind my readers that I posted my video review of the Augen Smartbook 6 months ago on January 29th as it's based on the Hivision PWS700CA and its cool RockChip ARM9 processor that runs Android in this video: http://armdevices.net/2010/01/29/android-laptop-review-hivision-pws700ca/
and that the Telechips ARM11 800mhz based Augen $149 7" Tablet that Engadget and plenty other blogs also are talking about seems to be based on the same 7" resistive tablet hardware design that I filmed 5 months ago presented by MAG Digital at CeBIT 2010 in this video: http://armdevices.net/2010/03/02/mag-digital-presents-windows-ce-that-looks-like-android-in-a-tablet/
To let you know my opinion. I think it is fantastic that Augen and KMart are promoting such cheaper Android Laptop and Tablet form factors as alternatives to the much more expensive Apple iPad and Intel Netbooks. Archos has also been selling the similarly priced Archos 7 Home Tablet on the worldwide market which I video reviewed 5 months ago, which is now broadly available in many retail and online stores below $200 for the 8GB version (and the 2GB version originally planned at $149, then $179 but for now they are mostly selling the 8GB version). That Rockchip based Laptop and Tablet platform also being upgraded to 1ghz still ARM9 to support newer Android 2.2 versions.
But as we have heard from Canonical developers and from hearsay and off camera chatter by Google people at the Google Q&A at Computex about Chrome OS on ARM Laptops, although the second generation 45nm ARM Cortex A8 with faster DDR RAM and faster I/O performance can be enough, the coming of ARM Cortex A9 platforms may be preferable to achieve the full desktop web browsing experience that most consumers may require for them to consider the ARM platforms as fully usable alternatives in the Intel/Microsoft dominated Laptop market. And the iPad and the whole bunch of smart phones that are currently spread all over the market, those may kind of set expectations at capacitive and ARM Cortex A8 performance at the minimum. So it will be interesting, capacitive touch screen manufacturers allowing, to see how soon and how cheap those capacitive Android tablet designs at full user interface speeds can reach the market. ARM9 and ARM11 resistive tablets are not bad for a start, they can give the consumers and bloggers a taste of what can be done with Android at retail prices below $200 and even below $100. The ultimate goal should be though that we should have full speed ARM Cortex versions of all these devices in all the stores, with the best capacitive screens for tablets or non-touch screens for Laptops, preferably Pixel Qi screens, and available below $200 without contracts, running free Linux based Android or Ubuntu OSes.
The new WiFi-only version of Kindle is $139, it uses the latest generation of E-Ink screens, with faster refresh, better contrast. It might be the worlds first e-reader to use the latest Freescale i.MX508 processor, which means the e-ink controller is integrated in the ARM processor SoC, which allows for 21% smaller design, 15% lower weight and significantly lower cost, while improving the processing speed at up to Cortex level to achieve faster refresh rates and optimizing more things such as doubling of battery runtime, faster e-book downloading times, usable web browsing speeds and more.
Sub-$140 connected e-ink e-readers are a big deal. Amazon is already selling more e-books to the Kindle e-readers than they are selling paper books. And consider that Amazon is the worlds largest online retailer of paper books. This e-book revolution has happened in less than 2 years since the release of the Kindle. Since Amazon's strategy is to make revenues and profits from sales of the content on their closed Kindle platform, it should be even possible for Amazon to further lower the price as needed, thus $99 Kindle shouldn't be far away. What happens, is that from about 5 million e-ink e-readers sold in 2009, there might be 15 million of those e-ink e-readers sold pretty soon.
Google is coming with the Google Editions E-book system later this summer, meaning next month or so. My big question is this, will Amazon allow Google's e-book system onto all Kindles through a firmware update? The way for Amazon to allow this to happen would be for Google Book Store to be only complementary to Amazon's current Book store. Basically, any titles that Amazon has in its store would be purchased through its own store, while google would only provide access to all contents that are not yet in the Kindle store. On those out of print or otherwise unavailable in Kindle Store contents, Amazon would be making a decent share of the revenue through a partnership agreement with Google. The point being that it could be great for Kindle and Google to have a partnership and a new firmware integration. If Amazon does not open its platform to Google, I expect we will see several new e-ink e-readers sold $99 or below from many manufacturers that will be using a special version of Android that Google must be working on to release with Google Editions. As Google will provide revenue sharing for the manufacturers of devices that access Google Editions and other online sources of monetizable contents, expect some type of AdSense for e-readers.
I am an Android fanboy. But I don't like the different custom Android user interface designs such as Motorola Blur, HTC Sense, Samsung TouchWiz, Sony Ericsson's nor Acer's custom user interfaces, I think they are confusing and they are like visual bloatware. Please investigate following:
1. Is there a home replacement in the Google Marketplace that returns the full UI in Android to the default Android UI in all of the different Android phones? If not, then why isn’t Google officially releasing this default Android UI in the Google Marketplace? Does anyone have any contacts at Google to whome they could ask about this?
2. Will that Default Android UI Home Replacement provide a way to have exactly the same UI as on the Nexus One? And without voiding any of the warranties? And without consuming any extra RAM memory or slowing anything down in any way?
It’s very simple, but we need simple and definite answer on this. And we need answers from Google officially. Thanks.
I do understand that the reason for each of these different layers of designs on top of Android is that each of the phone makers feel that they have to differentiate their Android offerings from the competition. That consumers have to think that they are buying a HTC phone and that only HTC phones can do that or look like that. And so on.
But I would rather that the Android ecosystem immitate the Windows world, let all the default desktop user interfaces look the same. Stop confusing the consumers. Let them recognize the true value of the whole Android ecosystem. And phone makers should preferably compete on hardware and features for the prices.
I understand this idea is disruptive to the current Android business plans of each of the companies. But please, could we at least just get a little basic Home Replacement somewhere deep in the Google Marketplace that offers anyone with any Android phone to set it back to the Nexus One styled basic Android UI design?
I've been told in certain comments that Launcher Pro and some other Home Replacements were good. But I am looking for a Home Replacement that offers exactly the same UI designs as on the Nexus One. Please let me know in the comments if you know of a near or fully Nexus One UI Home Replacement that is available for any of Samsung Galaxy S, Droid X, HTC Desire, Sony Ericsson X10 or any of the other Android phones, what your experience is with it and on which of the phones.
Having to void warranties on those $500 devices just to get a normal UI is a terrible thing.
Read more at droiddog.com: Hey manufacturers, leave Android alone!
The biggest threat to Microsoft's $62 Billion in yearly revenues and $24 Billion in yearly profits is the possibility that consumers and the enterprise start adopting sub-$200 Linux based ARM Cortex A9 laptops and desktops as the new standard for personal computing in the months to come. There is a high probability that Chrome OS and Ubuntu will turn out to work very smoothly on ARM Cortex A9 processors, so smoothly that most consumers might be satisfied with the experience of web browsing speed and for running basic applications like text editors but even basic video- and photo-editing once those are available on the cloud and powered by advanced HTML5 native code and caching mechanisms.
Microsoft clearly must be seeing this as the biggest threat to their core business and thus is probably preparing a version of Windows 7 for ARM. It won't run all the .exe files that run on Intel/AMD/VIA x86 processors. But there may be tools for developers to recompile the most important applications and to make new drivers. The challenge is for Microsoft to present such a lightweight version of Windows for ARM in a way that consumers will still pay for the Microsoft Windows OS experience even as prices of those ARM Powered laptops arrive at under $200. The profit margins will be low for Microsoft and this will require for them to implement totally different and bold business models if they want to try to keep the same numbers in yearly revenues and profits to not have their share holders sell their stocks. It would probably be based on Windows CE 7, but since that one looks much like Windows CE 6, it would be all about how they could upgrade the user interfaces to make it look and perform as much like Windows 7 as they can.
Some other possibilities as to what Microsoft may be doing with this ARM Licence:
- To release a new lower cost ARM Cortex A9 powered XboX to compete with the upcoming Google TV platform. 10 years ago, Microsoft launched MSN TV and for years there has been Windows Home Center Edition but it never really was a success like Google TV has the potential to be. Features need to be implemented in a cheap ARM Powered hardware such as HD quality video-on-demand streaming, casual and advanced 3D gaming, lean back web experiences.
- To release a Tablet centric OS to compete with Android and iOS, it would be related to Windows Phone 7 Series and with some features of the ARM version of Windows 7.
- Microsoft may want to design and control their own version of an ARM processor and keep it for their products. Like Apple keeps A4 for their products.
What do you think Microsoft is going to do with this ARM License?
This is not Google TV on ARM yet, but this is a major achievement already, Boxee can run the full Boxee software experience on an ARM Powered box. Boxee is considered to be one of the best user interfaces for media streamers and Web TV. I wonder how much would need to change in D-Link's Nvidia Tegra 2 Powered Boxee Box for them to be able to release a Google TV version of Boxee Box, and have the Boxee video navigation UI functionalities be an app on top of Google TV OS. My guess is the full Google TV experience requires HDMI input and output and an IR blaster (to integrate with existing cable/satellite boxes), thus would require a slightly upgraded version of this hardware.
For users who don't require the feature of old TV integration with cable/satellite set-top-boxes but who only want the future experience of VOD, media streaming and web TV stuff, it would be cool to know how likely or unlikely it might be to be able to load Google TV OS for ARM once it is open sourced by Google on this Box and have Boxee's complete set of features and user interfaces as a 3D accelerated app on top of Android. In my video interview with Boxee from CES, the Boxee representatives say that Boxee Box is designed to be open source and hackable:
There is an SD card slot on the side. We know that a lot of developers and hackers really like to side load the OS and have their own apps, so we are trying to make it as developer friendly as possible.