Feb 4th, 8:35am
Nicholas G. Carr, Author
Mr. Carr’s 2003 Harvard Business Review article and follow-on book, Does IT Matter, forced organizations to rethink IT’s role in developing and executing strategy. His current book is called The Big Switch examines the impact of cloud computing on business, culture and society. An excerpt:
At a conference in Paris during the summer of 2004, Apple introduced an updated version of its popular iMac computer. Since its debut in 1998, the iMac had always been distinguished by its unusual design, but the new model was particularly striking. It appeared to be nothing more than a flat-panel television, a rectangular screen encased in a thin block of white plastic and mounted on an aluminum pedestal. All the components of the computer itself – the chips, the drives, the cables, the connectors – were hidden behind the screen. The advertising tagline wittily anticipated the response of prospective buyers: “Where did the computer go?”
But the question was more than just a cute promotional pitch. It was, as well, a subtle acknowledgment that our longstanding idea of a computer is obsolete. While most of us continue to depend on personal computers both at home and in the office, we’re using them in a very different way than we used to. Instead of relying on data and software that reside inside our computers, inscribed on our private hard drives, we increasingly tap into data and software that stream through the public Internet. Our PCs are turning into terminals that draw most of their power and usefulness not from what’s inside them but from the network they’re hooked up to – and, in particular, from the other computers that are hooked up to that network.
This talk was an easy way to start the day. As I was absorbing my first caffeine infusion, it was satisfying to sit back and listen to Carr talk about how the early 20th industrialists switched from water power to electrical grids in the space of a few decades. The predominate power generation systems in the 1800s were privately owned water-wheels that drove mechanical factories. All this changed late in the century with the advent of centrally generated electricity. By 1930, 90% of all power generation came from grid utilities.
It is this type of disruptive technology event that Carr links to today’s increasing use of Internet computing resources, aka Cloud Computing. He makes the point that computers and storage that are privately owned are underutilized and require large labour efforts to maintain. Further, this labour is disproportionately spent on ‘keeping the lights on’ with only 30% of effort dedicated to creating new systems or directly supporting business automation.
Today there is a huge trend towards large-scale server farms that offer high efficiency and low cost. While this is nothing new — timeshare mainframes were all the rage in the 1970s — there are two significant differences with today’s Cloud Computing infrastructure:
- Server virtualization technologies offer enormous scalability and flexibility. Applications that need additional power can have capacity added almost instantly.
- Network capacity has finally caught up with computing power. For many years we have had enormous leaps in processing power that were not matched by increased network bandwidth. But today, high-bandwidth connections are the norm and we not longer need to be physically close to our computing resources to deliver enterprise applications.
Another important development is that the model is already proven with consumer (think Web 2.0) applications and massively successful business applications like SalesForce. The more enterprise IT sees these type of deployments, the more likely they will consider Cloud Computing as an option to reduce costs and increase capabilities.
Carr makes the assertion that Cloud Computing is a disruptive technology, and that users of traditional, self-hosted systems will likely be caught by surprise as this trend becomes more prevalent.
He does recognize that there are some important consequences of Cloud Computing:
- data is now centralized and connected over the web;
- information security is more important — mostly because the risks of breach and downtime are increased; and
- organizations may need to trade privacy (think Gmail) for convenience and low cost operations.
Governments have been slow to make the switch to Cloud Computing, likely for the three reasons highlighted above. There is a big concern around where data is stored — most Canadian organizations, especially government ones, will not accept personal or confidential information to be stored outside of Canada. Because Cloud Computing is so virtualized, it is possible that a service provider could not only store the databases in, say, the US, but move them to server facilities in one or more other countries without notice.
His final point was a slide showing a common power outlet surrounded by a large collection of now-common electrical appliances. Harkening back to his earlier point about electrical grids, Carr illustrates how ubiquitous electrical power spawned a wave of innovation in device design that would take advantage of that utility.
I think that Carr’s observations are well developed and he’s almost certainly correct is saying that many IT shops will move applications and infrastructure to the Cloud in the future. I’m less convinced that government bodies will be so enamoured with a technology that has yet to prove it is sufficiently secure, privacy-conscious or reliable as those system they operate today.
One thing is certain: the next few years of the disruptive technology that is Cloud Computing will be interesting to experience.