But I think I will prefer an Exynos 5250 for a top device. CPU performance will be similar (if not better, because Exynos use stock Cortex-A15) but GPU performance will be much better, and Mali T604 has unified shaders, able for GPGPU (OpenCL, DirectCompute, etc. …).

Hard times for Tegra and Atom processors.

I think Mediatek is already a success story. If you go to China, maybe more than 50% of the smartphones sold in that country are running on the MediaTek Processor, running Android smoothly. I think the MT6595 sounds awesome, it’s going to be amazing to see ARM Cortex-A9 and HSPA+ on sub-$100 phones, I expect them even to come with decent 4.3″ capacitive screens at those sub-$100 prices now.

I think all Android makers are doing good. And all will continue to do much better than they would without Android. The only reason HTC may loose market share would be if they are stubborn and don’t bring out cheaper phones but only want to focus on high-end. Then it’s normal, just as for Apple, that people only focused on high-end will loose market share. HTC must release Cortex-A9 performance 4″-4.3″ phones at below $200 if they want to win market share in this exploding worldwide smartphones market.

I think it will be very interesting to see what the Japanese Sony, Panasonic, Sharp, Fujitsu are going to do with Android these coming months. I think they can easily profit a lot from it. They can be leaders worldwide if they want. But it really depends on them agreeing to sell good cheap phones for pre-paid markets and not only expensive ones.

If the Huawei P1S is $299 or cheaper unlocked, I think Huawei can become one of the top selling worldwide smartphone makers. If they mass produce it at full speed, and really target $299 P1S and $249 P1, and $349 or $399 D1Q, Huawei is going to be awesome and probably my top recommended high-end phone maker.

TI will continaually do very well this year at Qualcomm’s expense.  Yes Mediatek, St Errisson, and Nvidia Icera have their SOC’s, but I you don’t see any widespread adoption yet at least until 2nd half of this year.

I still think that Windows on ARM is the dark horse.  Let’s face it, 3 months after Android’s ICS come out, you don’t see much ICS phones coming out.  Why?  Google would only work with only a few select phones.  You see software updates and android and iphone, but not much OS updates from android.  If Microsoft can streamline OS and driver updates for phones just like what they can do for pc’s, they can easily capture the Android and IOS markets.

It should be time for massive launch of LTE phones,now!

 i dont really like advocating http://www.phoronix-test-suite.com/ as the dev doesnt really add end user requests to the tests they want ,only listens to the commercial users,  but if someone techy with an existing rooted Tegra3 quad were to run the phoronix  tests then that would probably indicate Nvidia only use the cheaper slower ram , probably in the 6.4GB/s memory bandwidth or even less right now.

if Huawei can really claim theirs is 2x faster than a Tegra3, it wouldnt surprise me if they Simpy put faster 8.5 GB/s ram on their PCB to better match their SOC with other small tweaks.

 it’s always far easier to ask the company tech if their around what the real spec of that chip and but  etc is , so thats why i mention it in case you forget

also see http://eda360insider.wordpress.com/2012/02/02/3d-thursday-boosting-the-bandwidth-of-wide-io-sdram-to-1-tbitsec-through-standards-evolution/ 

as that “Wide” I/O Interface with Extreme Bandwidth” is getting some large standards backing talk now for the future.

”
3D Thursday: Boosting the bandwidth of Wide I/O SDRAM to 1 Tbit/sec through standards evolution
Posted on February 2, 2012 ”

“I want to focus on in this blog post is the issue of Wide I/O SDRAM. The JEDEC Wide I/O SDRAM spec (JESD229) calls for a 512-bit-wide data interface (split into four independent 128-bit channels) running at 266MHz with one transfer per clock cycle (single data rate signaling). 

This spec was created specifically for the low-power needs of mobile applications such as mobile phone handsets, where the total amount of energy available from a charged battery drives nearly all design decisions. (See “3D Week: JEDEC Wide I/O Memory spec cleared for use”) ….”

remember  Charbax the No.1 key question that needs answers this year is what speed ram/bus is in on all this new A9/A15 SOC review kit at Mobile World Congress

are they still restricting total overall throughput by using cheaper limited single channel DDR, at approximately 3.2GB/s max, or far better dual channel LP DDR2 or best LP DDR3, running at what exact ram speeds etc, at a top speed up to 6.4GB/s memory bandwidth, or better 2-channel DDR3 @532 MHz with up to 8.5 GB/s  or better ?

 did anyone in 2012 finally fit and use the with Samsung ”Wide” I/O Interface with Extreme Bandwidth
“The new 1Gb wide I/O mobile DRAM can transmit data at 12.8GB/s, which increases the bandwidth of mobile DDR DRAM (1.6GB/s) eightfold, while reducing power consumption by approximately 87% 
http://www.xbitlabs.com/news/memory/display/20110222201121_Samsung_Develops_Mobile_Memory_with_Wide_I_O_Interface_Extreme_Bandwidth.html

in other words how does their kit with their chosen ram speed match or better the generic PandaBoard ES running ARM On Ubuntu 12.04  runing the generic Ubuntu 12.04 development snapshot with the Linux 3.2 kernel and the http://www.phoronix-test-suite.com/ tests as the base line test today?
see http://www.phoronix.com/scan.php?page=article&item=ubuntu_1204_armfeb&num=1

If you want to hear my guesstimation, is that I think there is a good chance Archos is number 2 sub-$300 tablet maker in the USA, in front of Nook, Acer, Asus and any other pretending to sell sub-$300 tablets on the US market. It’s possible Archos is number 1 or close to number 1 for the $200-$300 or/and $200-$400 tablet market in the USA. On the other hand, the numbers show Archos US growth has not been as fast as it could have been, far from not as fast of a growth as in Europe and Asia, so it’s possible some Nook or some other price-dumped tablet alternative from Acer, HP, RIM or other briefly overtook Archos in that market, but for Q1 2012, Archos is well positioned to take a large chunk of that $100-$400 tablet market.

You are jus making assumptions, conjecture, and guestimating.

The source is Archos themselves, Archos knows the numbers of tablets and are likely to announce at their event on March 14th in Paris, where they also annonce profits and some other things about their strategy. Archos number of tablets sold and orders placed by resellers but not yet shipped (because of the processor delays and lack of cash to mass manufacture and ship before December 31st 2011 according to demand) may be about 2 million tablets for Archos in 2011. That’s double the amount of tablets Archos sold in 2010.

Archos quotes analysts such as GfK which is one of the top analysts firms counting consumer electronics sales in Europe. Samsung themsleves have said in the media that Archos is selling more tablets than them in France. Archos also quotes Gartner when talking about the 2012 prediction of the tablets market probably more than doubling compared to 2011.

Even at an average selling price of 100€ , it would only be 1,179,000 units for the entire year in Europe. That is not that impressive.

Assume all their income was from 400 euro tablets @
171,4 million euros yearly revenues the most optimistic figure is
428500 units sold worldwide which does NOT sound impressive.

Archos has always been the first to use the newest and fastest Texas Instruments processors and Archos has always provided those at the best price for value. TI has always been and always is on the bleeding edge of ARM technology.

Tegra3 is an excellent processor, but the marketing speak about it being twice as fast as dual-core is just plain wrong. You can always engineer benchmarks that may show it. But if you put the device in the hands of 99% of the consumer and measure actual load times and interface times of more than 95% of normal usage, you simply cannot see a difference. The Dual-core OMAP4470 at 1.8Ghz for example is definitely faster at nearly all use cases than the Tegra3 at 1.3Ghz.

Just think of ARM gpu Mali-processors and a lot of all types of radio standards like nfc,wifi,wiGiG
and hopefully USB Power Delivery 3.0 and in some years Thunderbolt USB.

Android at Home Hub from many developers and makers will make Apple TV obsolete!

Though I agree it is likely to take awhile but may be closer to 2 years.

Many of the decisions are being based on present limits of ARM hardware and don’t take into account all of the improvements that Cortex A15 is likely to bring.

Since one of the reasons to prevent emulation is because it would kill run time but if they improve efficiency enough for VM then they may change that consideration.

Their conditions to upload applications to the Windows Store, they are very restrictive. Something like a VM will not be accepted in Windows Store (like Apple also forbids), only Microsoft would be allowed to build one.

That’s very far from the full Windows experience that MSFT said. To me they are clearly creating their own wallet garden, cloning Apple. Windows on ARM is only a Windows Lite (or iWindows if you prefer).

Do you think Microsoft will allow a Virtual Machine Dalvik-like run-time to be approve through their app-store?

The only way to be cross-platform is to stick to something like WebOS Enyo and use a HTML5 framework. The only problem is the lack of performance/speed/native access to the underlining OS. Wait, that sounds almost like Metro apps.

If I have to rebuild from the ground my applications for Metro, there is no way I will switch to Visual Studio. My days of making Windows exclusive apps are over, and I will switch to whatever IDE that allow me to make crossplatform applications (my first thought is to use Delphi + Firemonkey so the switch from Delphi VCL will be easy, and I will be able to target in the near future Metro, Mac, iOS, Android, Linux, …).

I’m certain that I will not be the only one to switch to crossplatform if we have to confront the huge task to rebuild our applications to Metro (even the .NET developers must be very angry).

I think the big advantage of Windows will be gone with Windows 8. They lose all the huge base of legacy applications (they are forcing them to switch to Metro), and with the new applications becoming crossplatform the users remain free to choose whatever OS they prefer.

To answer this question we really need to ask will any Augmented Reality glasses be successful as a consumer product. I think it’s a no-brainer that they will be, the idea is just too powerful. The trick is to figure out when. If we are talking about …

The bigger question here is why Microsoft wants to push all ARM apps to these new APIs, when they are not putting these restrictions on x86 apps for Windows 8.  One possibility is that MS doesn’t really care about ARM and wants to slow it down, so whoever buys ARM tablets and laptops running Windows 8 gets pissed off and blames ARM, when they realize all their Windows apps won’t run on it.  The other possibility is that MS wants to take this chance to finally push everyone off win32 and make all apps go to WinRT and Metro, because any dev who wants their app to run on both ARM and x86 will have to port to WinRT to support ARM.  This is a big change and could be a suicidal risk: developers may decide they just don’t want to do all that work to go to WinRT and move to OSX/iOS or Android instead.

The key, as always, is going to be what the sales numbers for WOA and WOA apps look like, and it’s possible by then that the install base for iOS/Android, currently about a quarter of the Windows install base, gets big enough that devs really start defecting from the current market share leader.

There wont be “deepest hardware acceleration” He mentions “Soft drivers” It is the 1st bullet point in “Getting to the Start Screen section:Quote:This is a significant undertaking of very complex code since today’s GPUs are even more complex than the CPUs. To bring up Windows 8 on these new SoCs that did not yet have a graphics driver, and since ARM SoCs do not have the industry-standard VGA subsystem to fall back on for compatibility mode, our graphics team wrote a soft GPU driver that was capable of working directly against the hardware frame buffer. Aside from enabling development, it also enabled us to reimagine other things in Windows using the soft GPU driver when the normal GPU driver isn’t available.source: Steven Sinofsky I am starting to think you have a problem grasping the English language.

If your english is good enough, look up the word “germane.” How germane is cloud to the rebuttal against your Premier/Ulead Video suggestion? You can’t answer it so you move the goal post to appear that your arguments and logic hasn’t been cornered.

That is the archetype definition of “moving target”

From The MS blog:

Embedded MultiMediaCard storage (eMMC) is a de facto standard for storage on ARM devices (since most do not support SATA). This was an interesting challenge for us, since Windows expects a fast disk and very high bandwidth data transfer. In addition to supporting eMMC, we made several OS performance optimizations to reduce and coalesce storage I/O, resulting in fewer reads and writes to storage.
source: http://blogs.msdn.com/b/b8/archive/2012/02/09/building-windows-for-the-arm-processor-architecture.aspx

Good luck with a video editing system like Premiere/Ulead video to show up anytime in the near future.

App developers will be starting from scratch like they did with IOS and Android. 15 years of legacy Microsoft APIs have been abandoned. The new version of Visual Studio wont use those legacy APIs. 
It will be like the other hundreds of FTP apps on IOS and Android. It will be like the other hundreds of Office apps and Media apps on iOS and Android.They will all be re-written from scratch with the new framework.The thing is there is no “desktop app” advantage. There is no advantage with thousands of easily ported legacy apps that would make Win8 on ARM a killer platform.

I suggest you start reading MSDN, .NET and Silverlight developer forums before you talk on this subject. Go to Microsoft’s own blogs and forums. Download Visual Studio 2011/2012.

You are talking about Adobe, Google, Filezilla.

There wont be a Chrome or Firefox off the bat. It will have to be rewritten from ground up comparable to something liek Meego or WebOS. All start from day one w/ a level playing field.