I followed a meetup last night in Austin presented by Michael Cote of 451 Research and CloudCamp co-founder Dave Nielsen. John provided a review of the past five years of cloud computing and Dave talked about some current strategies in using PaaS. The overall program provided a nice and thorough review of where we’ve been and where we are.
The global nature of the cloud reaches to most nations. In the modern era, enterprise IT evolves at a similar pace worldwide, rather than in North America and Western Europe first, as in days past. Depending on what particular aspect you’re examining, it even moves more quickly in some developing nations – “eMoney” services on mobile phones in places such as Kenya and the Philippines are an example of this phenomenon.
Similarly, in the research we’ve been conducting at the Tau Institute since 2011 on the relative dynamics of IT development on a national basis, we’ve found far-flung bright spots in all regions of the world.
Cut Through It
With all this simultaneous development comes confusion. Cloud often continues to be defined in whatever manner a technology vendor wishes. Its perceived benefits become conflated, its power reduced.
In his talk last night, Michael Cote urged some renewed clarity by referring back to the definition of cloud published by the U.S. National Institute of Standards and Technology (NIST) in November 2011:
“Cloud computing is a model for enabling ubiquitous, convenient, on-demand network access to a shared pool of configurable computing resources (e.g., networks, servers, storage, applications, and services) that can be rapidly provisioned and released with minimal management effort or service provider interaction.”
The definition further deems the following characteristics to be “essential”: on-demand self-service, broad network access, resource pooling, rapid elasticity, and measured service. The NIST definition also lays out the familiar an still-used terms IaaS, PaaS, and SaaS, but defines public anad private cloud in ways different from today’s usage: private cloud is for the use of a single organization, and public for the use of many, regardless of whether the actual computing resources are on-site or provided by a third party.
The definition was written by Peter Mell and Timothy Grance, both of whom are still with NIST. Neither has a high profile, though are well-known and respected throughout the technology community. I don’t see either buying an NBA team or doing a GQ spread anytime soon.
I Prefer PaaS
Within this construct, it seems to me that PaaS is the catalyst in forming clouds out of the raw materials of infrastructure and ethereal functionality of software. Thus, Dave Nielsen’s look at PaaS served as a nice book-end to John’s presentation. Dave dives into the particulars of the Cloud Foundry PaaS and how to deploy an app with it.
So given these two talks, where are we going? Cloud computing, for all the good work that so many have done over the past few years, still commands only about 5% of enterprise IT spending.
Cloud revenues are counted in the billions – take IBM’s recently reported $7 billion in cloud revenue as an example – rather than the tens or hundreds of billions. IBM’s total, which includes such ancient technologies as Rational and Notes, accounts for only 7% of its total business. I imagine a similar percentage among its peers.
Amazon’s public cloud revenue appears to be running at a rate of about $4 billion per year. Its reent growth has tapered a bit, but is still running at about 40% annually. Amazon AWS has set the standard and is the clear market leader for public cloud. But even in this outlier case, cloud revenue accounts for only about 5% of the company’s revenue.
A negative person would ask, “what is wrong with the cloud?” A skeptical person would ask, “has the cloud been overhyped?” A more positive-minded person would say “we’ve only just begun.”