The Public Cloud
“Eventually everything will wind up in the public cloud.” Diane Greene, Google’s New Cloud Chief, May 6, 2016
“Private cloud is what most companies are adopting.” “No, public cloud is better and will dominate in the end.” “Nonsense, hybrid cloud is the way to go.” Need an answer? It’s in Manitoba!
CIOs report that private cloud is all the rage now. The Cisco’s of the world argue that hybrid cloud is the optimal path to adoption. Meanwhile, the big public cloud providers like Google and Amazon argue that it’s just a matter of time before most computing is done on public platforms like theirs. Each party with a dog in the fight proclaims they have the one true path.
Who is Right?
Why is this happening? Believe it or not, we have been to this movie before – more than 100 years ago when many companies had VP’s of Electricity. Are there lessons in it that will help us in our current controversy? Yes, but there are several nuances that also need to be incorporated and they come from our transitions to new forms of energy. Let’s explore.
Much has been written about how the cloud makes computing much more like a utility than it ever has been in the past. It’s ubiquitous, metered, easy to start and stop and increasingly inexpensive. One of the best books on this, published in 2009 – just as cloud really started gaining momentum – is “The Big Switch: Rewiring the World from Edison to Google” by Nicholas Carr.
Carr paints an easy to understand tale of how electricity first derived from individual power plants, then evolving into large centralized utilities parallels what we have seen in the provision of information technology. There we started with original walled gardens tended by elite acolytes and evolved to today’s cloud and the Consumerization of IT.
Carr also correctly points out some of the limits of the analogy but the lessons are nonetheless striking; especially, how each vested interest sought to advance their model while retarding the others. But in addition to the vested interests championing their preferences, there are fundamental economic forces that foster the private, public, and hybrid cloud frenzy.
Economics of Energy
Here is where we turn to the lessons of energy transformations and Manitoba. Vaclav Smil is a prolific writer and Professor Emeritus at the University of Manitoba. His research and explanations on the economics of energy provide the other part of the analogy to understand the emergence and transition to the cloud. While Smil is writing on the time it takes to move to renewable sources of energy, two of his observations help us understand why the many cloud deployment models are all present at the same time.
First, the amount of sheer growth required from providers just to meet the needs. Just like renewables have been around for a short time and despite accelerated adoption in many countries they still only meet a small fraction of total energy needs. Likewise, cloud too is relatively young (Salesforce.com was established in 1999 and Amazon Web Services in just 2006).
Gartner projects worldwide IT budgets for 2015 equal about $2.69 Trillion. The IaaS (Infrastructure as a Service) providers will do about $21+ Billion, or so, in revenue. Now, consider the implication. Cloud is certainly growing fast but in the big picture it is tiny and constitutes less than one percent (0.78%) of the total IT annual spend. Even when you toss in all the SaaS (Software as a Service) providers you are still only a drop in the bucket. There is an enormous build out required to meet just the current capacity need.
And that brings us to Smil’s other observation. The switch to different forms of energy took anywhere from 50 to 75 years because of the huge size and investment in the infrastructure. Likewise, electronic computing as we know it has been woven into business, government and academia since the 1950’s. Granted, a lot has changed and the technology refreshes itself ever faster but there are still a lot of investments in equipment and applications that need to be amortized. Not to mention old skills unlearned/new ones acquired. That sort of thing takes time – probably not 50 or 75 years but certainly not just 10 or 15.
Where does that leave you?
Be prepared for a longer haul. Private and Hybrid clouds are in many ways the logical transition deployment as the shift occurs to increasing Public cloud. Vaclav Smil leaves us with another point that we should keep in mind. These shifts in energy sources are daunting because they involve rejecting a way of life deeply ingrained. This will be no less true with cloud computing.
By John Pientka