You might think we make an awful lot of fuss about the cloud, considering that global public cloud spending amounts to a small fraction of IT spending overall — roughly $100 billion of a $2.4 trillion pie, according to IDC’s projections for 2016. But it’s all about momentum. IDC predicts public cloud spending will double by 2020, whereas overall IT spending will see something like a 13 percent increase by then.
More importantly, as Amazon, Microsoft, and Google keep demonstrating, the cloud has become the place to go for the most interesting new enterprise tech. Case in point: Check out Peter Wayner’s “10 new AWS tools redefining the cloud,” which walks through Amazon’s latest and greatest offerings.
Yes, there’s some cool machine learning and speech synthesis, but what strikes me most is Amazon’s increasingly granular understanding of app dev and deployment issues. Lambda@Edge, for example, distributes Node.js code the way a CDN distributes content to improve performance. X-Ray, which got a whoop from the crowd when it was announced at Amazon re:Invent, traces requests through complex distributed applications to isolate problems.
The incredible catalogue of AWS products and services, along with those offered by Microsoft Azure, Google Cloud Platform, and IBM Bluemix, has shifted the balance. For years I’ve been saying that open source is the engine of innovation, as startups “give away” inventive new software to gain market share. That may still be the case. But as InfoWorld’s Matt Asay has noted in several recent posts, the cloud is where the money based on that new functionality is being made.
Last week, for example, Matt asked the question “Who’s cashing in on containers?” and came back with the answer: the public cloud providers. Docker may have jump-started the container revolution, but most container deployments are in the cloud, and the public cloud services to manage containers are racking up the lion’s share of the revenue, not Docker the company.
Matt applies the same argument to platform as a service, open source databases, and big data — increasing proportions of the revenue for all of them are accruing to the public cloud. Standing up new software and infrastructure on-premises seems less attractive than it once did, particularly when you consider that obsolescence looms as soon as you unbox that new technology. In the public cloud, on the other hand, you can get the benefit of improvements as soon as they’re rolled out.
Even VMware, whose SDDC (software-defined data center) stack actually delivers a viable private cloud, struck a deal with the devil last year with VMware Cloud on AWS. The resulting hybrid offering will make VMware private clouds more elastic, but once those workloads extend to AWS you have to wonder whether VMware can retain control.
Andrew Oliver recently provided a great rundown of what you should do to prepare for the cloud in “9 steps to cloud readiness,” which concludes with the most important point of all: Start thinking cloud first. It will be years before most enterprises initiate mass migrations of workloads to the cloud, but get your mindset straight now.