Microsoft versus Intel: No matter who loses, we win dows.

Microsoft already has a huge database of .exe for monitoring and managing compatibility, it’s called Microsoft Application Compatibility Toolkit http://www.microsoft.com/downloads/en/details.aspx?familyid=24da89e9-b581-47b0-b45e-492dd6da2971&displaylang=en

Yes I believe Intel’s move on Moblin/Meego is what convinced Microsoft to put their 1000 top engineers to destroying Intel by making sure Windows 8 is fully ARM optimized, and that Windows 8 also brings desktop/laptop computing to the type of UI that can work on Tablets, Smartphones even (so Windows 8 on ARM will be quite close even fully compatible with Windows Phone 8).

I thought my Archos would be fast enough. It’s like saying they will be here soon enough. That’s not soon enough! 😉

I’m with Richard Stallman on .NYET. I think it is a Linux Trojan coded by a guy who is anything but objective. We were told that it would be taking over the net and way to cool to resist… years ago. It is currently a marginal technology and next to other technologies out there really deserves to remain so… Unless we want to risk something like the GIF patent fiasco on a much grander scale. I’m simply doing all I can to remain immune and there or lots of options.

First of all Intel may have a point, but I can’t say for sure. In theory if done right and everything is making clean Windows calls than as a matter of calling the API you could run any Windows program on ARM, assuming Microsoft took that tack and making one more huge assumption. That assumption would be whether there is any assembly code in the program. That means games and legacy apps optimized for performance on x86. Alternately if Microsoft takes the easier course it means every program needs a recompile. Knowing Microsoft and that they could not even make their latest mobile phone platform backwards compatible this has some probability. Remember, Microsoft is notorious for promising big things and delivering small internal changes and some cosmetics. That’s not going to work with with a new platform. Last time Microsoft  tried to build for RISC was Intel’s Itanium (Itanic). After years of getting nowhere it flopped and AMD launched X86_64 leaving Intel with it’s pants down. It took years before Microsoft had an x86_64 release of Windows out of beta.

Clearly Intel has a motivation to spread FUD on Windows on ARM. However remember Intel is also working on MeeGo as they are half the proposition having contributed their Mobiln OS which they spent years on. If I had to lean one direction or the other I tend to trust Intel a little more. Additionally your suggestion that Microsoft could just recompile old Windows programs has a real issue. Yes, they could be motivated to get product on ARM and they could help developers. No they are not going to have access to all the source code and time for lots of debugging on a large scale.

Again Microsoft is notorious for breaking promises on releases. Time will tell. Intel may have had a motivation to discolor things, but so does Microsoft and the proof will be in the pudding. It’s not so simple.

Agreed, to say there are 4 different versions of ARM is like to say there are 10+ different versions of x86, pentium II, pentium II, pentium IV, core2duo, core series, AMD’s, VIA’s, etc…  I’m sure that Microsoft will write up some drivers that will optimize for a certain CPU, but in terms of software, one software will be written for ARM.  Microsoft will be forcing the ARM cpu manufacturers to comply to their standards.

The idea that there will be 4 different versions of Windows for ARM is also quite silly.  There are as many ‘versions’ of windows as there are PC models because the Windows installation program ‘builds’ each installation according to the hardware it finds.  The next generation of Intel Atom will be an SOC and it will be a different SOC than the AMD Bobcat.

The App store you propose has two big advantages, in my opinion.  First, you can weed out all the rubbish legacy apps.  Second, you have much greater immunity to viruses if you use only the software in the app store because only secure programs will get in there.

On the plus side, when windows 8 ARM come out, we would see alot of cheap Windows 8 powered media players and Wyse terminals.

What I think has been told in some of my videos is yes ARM Cortex-A15 starts sampling perhaps end of this year or early next year, but until the processors are ready for mass production and to be included in actual products there always needs to be a bunch more months.

On the other hand, companies like Nvidia and probably TI, Qualcomm and others are also investing huge amounts of money in R&D to speed up their work to the maximum.

 Hi. I thought that OMAP 5 (Dual Core A15 + Dual Core Cortex M4) will be in production at the end of this very same year, and in devices in 2012.

This processor can match Windows 8 schedule.

Didn’t they announce no effort at all for x86 compatibillity just a few days ago? It’s a HUGE job (to get good performance) anyway. I guess by just translating operating system calls you could make 80% of the apps fast enough though.

You’re right that .NET could become their “one tool to bind them all”, but at present it’s absolutely not. I’ve seen a couple of apps adding .NET as a dependency recently but pure apps? No way! Try to name 5 well known desktop apps written entirely in .NET. Actually I can’t even think of one (even visual studio 2010 has huge parts written in c++).

Cortex-A15 prototypes will be shown early next year for commercial release by the end of next year, though there are also very fast Dual-core 1.6Ghz and Quad-core Cortex-A9 being released sooner. I think those are fast enough.