I'm in Korea this week to keynote at the IDG "Virtualization in Action" conference, and I observed a fascinating disconnect while here.
First, and this should be no surprise, the conference could just as easily be titled "Cloud Computing in Action." (Indeed, my speech was titled "Creating a CIO Cloud Computing Action Plan"). The fact is, most conversations about server virtualization these days turn into cloud conversations -- despite the fact that most surveys show that typical IT organizations are no more than 40% virtualized inside the data center. I've long said that "virtualization is not a product, it's a journey." By this I mean that once one has begun using virtualization, the additional functionality available by taking the next step of the journey is usually irresistible. And the step after that.
Consequently, most conversations about server virtualization these days inevitably lead to cloud computing, and this conference is no different. And that's what leads to the disconnect.
However, before I jump into that, I want to share a fascinating anecdote we heard from another speaker, Raj Dhingra, GM of Desktop Virtualization for Citrix. As part of his discussion about the move to computing on a variety of devices, he shared this story about the iPad: in an unnamed Silicon Valley chip company, someone demo'd an iPad to the company's CIO. A day later he bought an iPad and began to do all of his work on it. He showed it to the CEO and a few days later the CEO had an iPad and began to do all *his* work on it. A few weeks later and the CEO has arranged for all 6,000 employees of the company to get iPads!
Cloud Just One of Several Revolutions
This anecdote aligns well with the trend we discuss with our clients, which is the fracturing of the device universe. The old desktop/laptop model is splintering, with those devices being joined by smartphone and now tablets. The thing is, it's not going to stop there. The spread of computing into devices has only started, and soon computing will be performed by things we will barely consider to qualify as a device -- not to mention sensors and actuators.
The general theme to be picked up by this anecdote is that the relatively well-ordered computing universe of a few years ago is being rapidly transformed by a revolution in client devices, which is being matched by a revolution in computing services, as traditional data center computing is being supplemented (or, depending upon the individual IT organization, displaced) by cloud computing.
Which is why I was surprised at this conference, which, as I noted, ended up being a cloud conference rather than a virtualization conference. Another speaker, Jim Fagan, Managing Director of APAC for Rackspace, did a survey of conference attendees (˜ 220 people) of their use of cloud computing.
When queried about who was doing a private cloud implementation, not a single hand went up. When asked who was using external public clouds, no more than two or three hands went up. And when the question of use of SaaS was posed, only a couple of people raised their hands.
While some of the small response might be attributed to Jim's presentation being directly after lunch, it seems that cloud computing here in Korea is today being seen as a research topic rather than an action item.
So the disconnect for me was the conference's total focus on cloud computing, and the apparent total lack of cloud adoption on the part of attendees. I'm not sure how to resolve the disconnect, but here are some thoughts about what it means:
• The U.S. is actually further ahead than one might think. There's lots of discussion about how U.S. IT organizations are being very deliberate (not to say, deliberately slow) about moving to cloud computing, but I guarantee if Jim Fagan had posed his questions to an audience of U.S. CIOs, there would have been a bunch more hands in the air. So perhaps U.S. CIOs are really more adventurous than characterized in the popular imagination.
• Korean IT is extremely conservative. This possibility was proposed to me by a Korean IT director, who said that IT organizations in Korea move very, very slowly to new trends in IT, whether they supposedly deliver great value or not. Even if they do, Korean CIOs do not get overly enthusiastic about new trends, preferring to wait for competitor companies to blaze a trail. When everyone is waiting for someone else to do the trailblazing, it results in a lot of hanging back, waiting for another IT organization to be brave.
• Cloud computing is really just an overblown vendor initiative. Just as car manufacturers rev up the publicity machine every September, this perspective views cloud computing as just another attempt by tech vendors to excite customers into another refresh cycle.
I'm not sure if any of these items is false, and all of them may be a little bit true. Overall, though, this disconnect points out certain facts:
• CIOs are confused about cloud computing. It's a different confusion than last year. Then, CIOs were confused about exactly what cloud computing is. Today, CIOs have a better sense of what cloud computing is. Now, they're confused about how to apply it in their environment. Frankly, most vendors aren't helping here. Every tech company around is telling CIOs that the road to the promised land of cloud computing goes through a purchase of its product, and I've detected in a number of CIOs the very real sense of vendor overload. As marketing guru Theodore Levitt once observed "People don't want to buy a quarter-inch drill, they want a quarter-inch hole." CIOs want the benefits of cloud computing, not the feeling that they're helping a vendor makes its numbers. Given the din of vendor overload, CIOs are retreating into full paralysis mode, reluctant to make a move until the dust settles.
• It's early, but you'd better prepare for cloud computing. Even if you think cloud computing is overblown, there's no denying that the benefits of even the early stage server consolidation version of virtualization offers enormous financial benefits. The first step toward cloud computing is having a fully virtualized infrastructure, and there's no reason (or excuse) not to fully implement server virtualization. In other words, even if your long-term direction doesn't currently include cloud computing, the cost advantages of virtualization are such that you *must* pursue the highest possible virtualization adoption possible in your environment. Getting stuck at 40% isn't a shortcoming, it's a disaster.
• Standing pat risks being left behind. Even though the vendor overload results in quite sensible delay, many companies are moving forward. It's uncomfortable, it's frustrating, but some companies are nevertheless implementing cloud solutions. If one of the early adopters is a competitor of *yours*, you face the very real possibility that waiting for the dust to settle may leave you trailing at the back of the pack. We encourage companies to implement small scale cloud applications with a view to gaining experience that can inform subsequent decisions. Better time be spent learning rather than waiting for clarity in a movement that may not become clear for several years, by which time a permanent gap between early and late adopters may be in place.
So even if the disconnect described at the beginning of this post is true, the issue is one of timing, not of existence. In other words, the issue is a when, not an if question. It's going to be interesting watching the wave of cloud computing adoption wash around the world.
Bernard Golden is CEO of consulting firm HyperStratus, which specializes in virtualization, cloud computing and related issues. He is also the author of "Virtualization for Dummies," the best-selling book on virtualization to date.