At MIT’s Emerging Tech conference last week, I listened to frighteningly smart people debate the future of this old-is-new technology concept that we call “the cloud.” Microsoft showed up too, to share its vision for cloud computing. Memo to Microsoft: From what I’ve heard, you don’t know cloud.
I looked at your early vision for the cloud, and frankly, I think IT veterans will see right through it.
Unfortunately for IT leaders and users, we are still very much in the land grab phase of cloud technology (You might say sky grab, but I have put a firm personal moratorium on cloud puns.) Almost every technology vendor is loudly trying to articulate its definition of cloud computing and strategy for it. This group ranges from VMware, with its elaborate plans for a cloud partner ecosystem unveiled at the recent VMworld show, to Microsoft, with a talk given at the MIT conference by Craig Mundie, Microsoft’s chief research and strategy officer.
I am convinced Microsoft has the most to lose if the vision of cloud computing as an on-demand extra data center and go-anywhere user office comes to pass.
I am also convinced that Microsoft continues to ignore some of the reasons why end users and IT leaders are drawn to the idea of Web-based computing and the cloud in the first place.
VMware, Google, Amazon and a slew of others are working diligently to tap into the millions that can be made from enterprise IT departments who want to simply not deal with so much hardware infrastructure and software maintenance, or contract with cloud providers to tap into extra compute and storage capacity on demand.
When enterprises can be convinced of cloud computing on the availability and security fronts, which are both still very much under scrutiny today, we will see a select group of “super utility” sized companies dominate the cloud, agreed a group of panelists at the MIT conference, including VMware co-founder Mendel Rosenblum (fresh from his public departure from the virtualization giant); Amazon’s CTO and VP Werner Vogels; and Salesforce.com’s executive VP of technology, Parker Harris.
It’s a when question, not an if question, whether this select group of super-big cloud providers becomes viable enough for enterprises, Rosenblum told the MIT crowd.
“The question is how far off is that,” he said, noting that while the death of the IBM mainframe has been announced many times, on it ticks in the backrooms of many enterprises.
One really interesting question for both enterprises and consumers, of course, is what is the future of the operating system as we know it today, if the computing model changes to enterprises running even some, not all, of their work via the cloud, and consumers using mostly Web apps and apps and services in the cloud.
“The OS is going to go from thing that was really crucial…to this thing that the application developer chooses,” Rosenblum said. OSes aren’t going away, he noted, but they’re not going to be front and center for users. When, he asked, do users ask about items like BIOS anymore? Virtualization, one of Rosenblum’s passions of course, has long abstracted hardware from the OS.
Assuming the industry can handle the security issues, the combination of virtualization and viable cloud computing providers could, in a few years, create a Texas-sized neon exit sign for consumers and enterprises dying to leave Microsoft operating systems behind.
And between Linux and Apple, non-Microsoft OSes will continue to flourish.
On the enterprise IT side, one reason that IT leaders get excited about the cloud is that the model could free them from the giant sucking sound currently heard coming from Microsoft licensing fees on the server and client sides.
Then there’s Windows PC support and maintenance costs. After decades of work to simplify PC management, the care and feeding of Microsoft-based machines still costs corporate IT much too dearly.
So what did Microsoft’s Mundie have to say to the MIT conference attendees about Microsoft and its plans for the cloud? He pitched a vision of a “client/cloud” strategy, within which a client OS still carries a good deal of weight for computing tasks.
“You have to have a to-and-fro between local and centralized data services,” he told conference attendees.
Note to Microsoft: Google, Amazon, VMware and your other rivals are betting that enterprise IT leaders and consumers will have the same reaction to that line of argument that I did: C’mon.
Google is well on its way to showing that the vision for cloud does not require a big, expensive, high-maintenance client-side OS.
Mundie also discussed an early vision of what Microsoft calls “spatial computing,” including 3D virtual worlds, speech and touch interfaces, and a combination of work being done by the cloud and client computer and OS. It’s visually interesting and utterly theoretical at this point.
Microsoft has to think big right now, on many fronts. But based on what I heard earlier this month at VMworld and last week at MIT, Microsoft’s rivals are thinking smarter about the cloud.
Microsoft has to fight for its client OS technology, that’s a given. But it’s going to be harder than ever for them to win that fight with IT veterans and consumers who’ve heard the Microsoft promises before.
Shoes don’t stretch. Bad romances don’t change. And Microsoft doesn’t make svelte operating systems, for use in the cloud or elsewhere.