ARM Powered Unified Windows: Opportunity, Risk, and Reward ...
Zax,
<< Times will be very exciting when Windows "8" comes out. If history repeats itself, I am pretty certain that Win 8 will catch some initial malignment like Vista did and all will be forgiven in a successor that will be more widely adopted. >>
I feel the same way, both in terms of potential initial 'malignment' and a more widely adopted, smoother running successor. Stripping back the WinMob CE kernel to bare bones, introducing a code break, and fleshing out the new write to create WinPhone OS 7.0 was bold and courageous.
Nokia flubbed the dub by not doing similar with Symbian despite the pain it would have caused. They did a fine job moving Symbian to open source and getting over the IP humps but got bogged down with code development. Too many decentralized and competing teams were involved. Microsoft and its early adopter OEM licensees underwent pain with the rather simple NoDo upgrade to 7.1, but experienced some positive learning in the process. The 7.1 --> 7.5 Mango update went remarkably well, however, and even Samsung who really struggled with the NoDo upgrade got through it unscarred although slower than HTC and others.
Despite the fact that WinPhone OS 7.0 was sadly lacking in basic modern day smartphone features and functionality the Metro UI justifiably got very positive critical reviews in its early stages. Mango closed the gap with iPhone and Android considerably but its comparative shortcomings remain(ed) several and obvious.
None of us should underestimate the challenge Microsoft faces in moving to a unified Windows for Desktop, Notebooks, Ultrabooks, MIDs, and Smartphones (Mobile Multimedia Computers as Anssi Vanjoki called them) thanks to a common kernel with shared underlying code, components, and user experiences. The upside potential opportunity of doing so is huge but the attendant downside risk looms large.
<< Much of the initial issues with Vista were the software catching up to the newer model and tweaked to work well within it, and the rougher edges of any new model getting smoothly polished. I can't imagine Windows "8" not experiencing similar initial difficulties. >>
The OEM licensees and component and peripheral manufacturers struggled with the newer model as well. Drivers were a real issue. Much smoother with 7. While I skipped Vista (other than babysitting and troubleshooting systems purchased by friends) and stuck with XP, I did prep a new highly discounted Vista SP1 equipped system in early October 2009 for the free upgrade to Win7 I installed a month later. Fortunately although Vista's SP2 was not yet in auto download distribution it was available for download. I had one serious conflict with a software app (a files and folder manager replacing Explore) and Vista SP2, and after the update to Win 7 I waited about 6 months for Canon to update drivers for software to manage software for my digicams. Other than that t'was a breeze and Win7 was everything Vista had promised to be and more.
I can't really imagine Windows "8" not experiencing similar initial difficulties either, but the ability of working through it relatively rapidly is going to be critical. Microsoft doesn't need another Vista and WinPhone really can't afford to get bogged down. For all its promise WinPhone must start to gain legitimate traction with Mango/Tango in the coming quarters and then get Apollo to market relatively cleanly to keep that momentum going and legitimately become the 3rd smartphone ecosystem.
<< I am certain that Windows 8 on x86 will do great, but the real question is whether Windows 8 on ARM will compete well. >>
That IS most certainly a legitimate question. I have great respect for ARM Holdings, and for Qualcomm who I think will be considerably more focused on ARM architecture for Windows than Intel or AMD, but while their challenge is considerable I think they'll work through it well.
One advantage I see for Microsoft in the transition is that the WinPhone ecosystem of OEMs and silicon supply is comparatively small compared to Android, but the key primary dedicated strategic partners (Qualcomm and Nokia along with ARM) are well chosen, sharply focused, and highly motivated. Qualcomm and Nokia San Diego are in close physical proximity (as are their teams in Bejing) with Microsoft not too far north, Despite the fact that Nokia's Salo and Tampere Finland teams are physically remote the partners are making diligent and frequent use of HP's Halo video conferencing technology ...
h71028.www7.hp.com
Halo, which [Microsoft's Terry Myerson} said he hadn’t used before the Nokia deal came together. The [video conferencing] customized room lets a team in one place seem like they are separated only by a window from colleagues sitting halfway around the world. “It’s as if we are sitting in the room with people in Finland or London,” Myerson said. “It’s like something out of ‘Star Trek.’” Halo is used, on average, five or 10 times per week to bridge the engineering teams from Nokia and Microsoft.
There's also been important cross-fertilization of key personnel ...
But sometimes that’s not enough, of course. Nokia has transferred a top executive — Waldemar Sakalus — to Seattle to oversee the Microsoft relationship, and is spreading hardware development work across several locations, including San Diego, Calif., and Beijing, as well as two sites in Finland: Salo and Tampere. ... Nokia also hired Kevin Shields, a former member of Myerson’s Windows Phone team, to oversee Nokia’s efforts to build on top of Microsoft’s operating system.
[The quotes above are from an interesting July article by WSJ's All Things D writer, Ina Fried (formerly Ima Fried. See the post that follows for the article and another from last November on the Qualcomm connection]
<< I wonder if Windows on ARM will have a chance at getting any real penetration outside of Windows Phone and Embedded platforms. I haven't heard this discussed much... but it seems a fundamental question and challenge. >>
Most likely in the initial years the Win on ARM focus will be on WinPhone and tablet slates. I've wondered recently to what degree ARM will power Ultrabooks or hybrid laptops or convertible Ultrabooks with detachable tabletlike displays in the relatively early going. Intel architected Ivybridge or Fusion Trinity powered units are quite likely to have the strong upperhand in that arena for some time. I can't yet even imagine using an ARM powered desktop.
<< What I most look forward to seeing is how Visual Studio will target the Windows 8 RT. >>
I really haven't figured out or even paid much attention to how the requisite dev tools will evolve or what the fate of Silverlight will be.
<< I do so enjoy trolling the nutball fringe. Did you ever step on fire anthills when you were a kid? ;) >>
I have an Evil Twin to troll and agitate the nutball fringes. As for fireant hills ... no, not growing up in New England, but I sure did as an adult playing golf on unfamiliar courses on outings in the south. <g> ###
Cheers,
- Eric - |
