I loved this quote by George Gilder in Forbes.com’s Coming Creativity Boom, Referring to Cloud Computing, Gilder states, “The rule for the new architecture is that hardware softens on the edge and software hardens at the core.” For those who have tracked my uses of the terms “core” and “edge” in this blog over the last year, you will recognize why this quote so resonated with me. For those that are not familiar with my use of those terms, please check out here and here, for examples.
I’ve referred to Cloud Computing a few times in this blog (who hasn’t), mostly in a tone that has been “pro” this important trend. Lately, however, I’ve seen more negative commentary – more disbelievers and naysayers highlighting the reasons why Cloud Computing is not for all, or why it’s dangerous, or over-hyped, and so on. Yes, the cautions are appropriate – like any paradigm shift, the technologies behind Cloud Computing are immature, our enterprise infrastructures for working in this new paradigm lag the new arhcutectures, and there are real and valid concerns about security, reliability, and serviceability.
However, I’m frankly puzzled that in October 2008, there is as much naysaying as there is! From my perspective as one that has studied trends and patterns in enterprise computing for some 40 years, several truths seem to me, to coin a term, to be “self-evident.” These include:
- In many respects, we’ve used Cloud Computing principles before – quite successfully. In the 1960′s and 1970′s, service bureaus were an important part of the enterprise computing scene, especially as they evolved beyond overnight batch computing to the provisioning of on-line computer terminals. These often supported innovative and extremely powerful forms of end user computing (what some of us refer to nowadays as “edge computing”) such as complex modeling and simulation, and data analytics. In fact, many of these tasks could only be done through remote service bureaus – they had the computing power and the sophisticated software that the typical enterprise IT set up did not.
- Many of the services to which users need access – processing cycles, data storage – truly are commodities. When communication bandwidth was limited and/or was expensive, we needed those processing cycles and data storage to be as physically close to the user as possible. Today, as we approach 10 gigabits and more as the lower threshold in many parts of the world, this computing power can be anywhere – just as long as we have ubiquitous access to high bandwidth. Yes, I know we are not there yet, but the rapid growth in high bandwidth telecommunications is a reality, and shows no sign of slowing down. For the skeptics, you might dig into Gilder’s Law and how that is playing out in reality.
- Energy costs and climate change are impacting decisions about where to place data centers. Compounding these issues is the need to make IT costs more elastic – to be able to react on demand to both increases and decreases in compute power. Cloud computing satisfies this increasingly valuable requirement.
- We can do things in the Cloud that cannot easily be done in the enterprise computing environment. Just as all those teletypes and green screens hooked up to service bureaus gave us access to powerful modeling languages such as APL, analytics software such as SAS and SPSS, and to the so-called Fourth Generation Languages like Focus and Nomad, so too will Cloud Computing give us access to capabilities for business innovation, and for participating in communities in ways that cannot be achieved within the enterprise firewalls.
- The Cloud is inherently collaborative and integrating. The standards that make Cloud Computing feasible, also make it easy to “mash up” new capabilities, and to incorporate a geographically and thematically dispersed communities (customers, suppliers, interest groups, and so on). By contrast, the vast majority of enterprise computing that has been developed over the last 20 years or so is inflexible, non-integrated and difficult or frequently impossible for users who were not part of the original design intent, or who are outside of the enterprise firewalls, to take advantage of.
- In a nod to the economic climate, I believe Cloud Computing is more cost effective – certainly, delivers a more flexible and palatable pricing model – one that is inherently value-based (pay “by the drink.)
My recommendation to IT leaders is this – stop reading about and worrying about all the reasons Cloud Computing might not work for you, and start identifying situations where it could work – and where it might in fact give you a business edge. OK, to be realistic, do continue reading what the smarter and independent analysts are saying about pros and cons – you have to be educated on this rapidly evolving field, but my point is, find ways to make this work where it makes sense. Experiment – one of the attributes of Cloud Computing is the ease with which you can get into it and play. Don’t think in terms of a wholesale switch – you don’t need that, and it probably does not make sense for most of us today. (Though if I were counseling a start-up, I’d think seriously about leveraging the Cloud to the full!) Again, think of the “edgy” things you’d like to do for or with your business partners, and approach those as your learning opportunities for this new architecture.