In 2008 I gave a talk at a conference on The Promise and Reality of Cloud Computing. In his closing remarks, the conference organizer noted that most everyone had agreed that something big and profound was going on, but they weren’t quite sure what it was they were excited about. “There is a clear consensus that there is no real consensus on what cloud computing is,” he said.
At the time, most people associated cloud with utility computing and IT-as-a-Service. Technology writer Nicholas Carr was also a speaker at the conference. In his talk, based on his then recently published book The Big Switch, Carr nicely framed the historical shift to cloud computing, comparing the evolution of IT to that of power plants. In the early days, companies usually generated their own power with steam engines and dynamos. But with the rise of highly sophisticated, professionally run electric utilities, companies stopped generating their own power and plugged into the newly built electric grid.
IT was then undergoing a similar transformation, as cloud service providers were starting to offer IT-as-a-Service with near unlimited scalability at very attractive prices, based on a flexible pay-as-you-go delivery model. Early cloud adopters were mostly focused on improving the economics of IT, giving them the ability to launch new offerings or expand their current ones without major investments in additional IT infrastructure while paying for whatever cloud resources they actually used.
Cloud continued to evolve and advance over the ensuing years. As an article noted about a 2013 cloud conference: “There wasn’t a single speaker who started off a session by saying, ‘Let's define cloud computing.’ That gets tiresome when seen in session after session, year after year, so its absence is gratefully received. This is a clear indication that the industry has moved beyond elementary knowledge-gathering and onto the practicalities associated with cloud implementation and rollout.” Still, debates continued on the true definition of cloud computing. Do private and hybrid clouds qualify or only public ones?
“Rather than sticking with a single cloud-computing service, a growing number of companies are working with multiple cloud vendors, spurring a wave of tech partnerships and new tools aimed at helping them integrate and manage diverse cloud-based infrastructure and software. The strategy, which is spreading despite added costs and headaches, in part reflects a shifting view of the cloud from a more efficient, cost-cutting tool, to a source of business innovation and new revenue, chief information officers and industry analysts say.”
Surveys indicated that midsize and large companies are currently using an average of 8 different cloud providers for various applications and services, and the number is expected to rise to 11 over the next two years. Integrating these various applications and services from different cloud suppliers has become a major issue for companies, leading to the emergence of multi-cloud management platforms, such as IBM’s Cloud Automation Manager.
This marks the evolution of cloud computing from its early IT-as-a-Service days to now becoming a major part of the digital supply chains that most companies rely on. Let me explain.
To me, cloud computing is essentially the Internet of services. Over the past decade, computers and data centers have been disappearing into the cloud. It’s what has always happened when technologies become ubiquitous in society, - electricity, telephones, television, - and now IT. While 10s of billion of devices are now everywhere, - growing to 100s of billions over the years, - their back-ends are nowhere to be seen. They are quietly doing their highly disciplined work of generating and transmitting their services and information to users and IoT devices out there.
Cloud computing represents the industrialization of the data center. Many data centers evolved over the years with limited architectural discipline or company-wide governance. IT organizations often spent the bulk of their energies on the maintenance and integration of their legacy applications, some having been developed by different departments within the business and some having been inherited through mergers and acquisitions. Not surprisingly, these older companies were then challenged to respond to fast changing market conditions, especially when it came to new customer-facing applications, which generally require massive scalability, flexibility and agility.
IT had to become much more disciplined in every aspect of its operations. Data centers have now become the production plants of cloud-based services, requiring major advances in productivity and quality to be able to support the exploding demand for mass customized applications, information and services of all kinds. This has forced companies to rely on cloud-based digital supply chain partners, - both for standard, commoditized services and for leading-edge services requiring highly specialized skills, - so they can better focus their organizations on those in-house services that provide truly differentiated business value. Some companies will choose to become service providers themselves, by selling their business services, - which they previously used only internally, - to the general market, thus developing new revenue sources.
Thirty years ago, we saw something similar happen in manufacturing. Before that time, most manufacturing plants were fairly inefficient by almost any measure, and were turning out products of varying quality. Then, driven by the huge success of Toyota and other companies around the world, the industrial sector and academia discovered the merits of applying engineering discipline as well as holistic, systems-wide approaches to manufacturing, leading to major improvements in their productivity and quality. At the same time, the costs of managing complex, geographically dispersed supply chains started to come down, - largely driven by global, integrated IT systems and sophisticated supply chain management applications. Companies started to rely on supply chain partners for many of the components once built in-house, enabling them to better focus on their key core competencies.
Most companies have adopted a hybrid manufacturing model, relying on the supply chain for many of their components, while continuing their in-house manufacturing, in particular, the integration, testing and distribution of their final products. The IT industry has been embracing a similar hybrid model.
In a 2015 special report, CIOs Say Hybrid Cloud Takes Off, CIO Journal editor Steve Rosenbush writes that CIOs “are knitting together a new IT architecture that comprises the latest in public cloud services with the best of their own private data centers and partially shared tech resources… For now, adoption of the hybrid cloud is motivated by the need for improved collaboration and greater flexibility and efficiency… CIOs are demanding a way to combine the best of the cloud with their own localized data centers. Few companies or organizations are willing or able to move all of their IT to the public cloud, yet most of them are eager to move past the anachronism of the isolated, on-premise data center.”
In the end, cloud computing is such an important and interesting subject because it lies at the intersection of three major 21st century trends: the evolution of the IT industry; the rise of digital supply chains; and the growing importance of services in the digital economy.
While cloud computing has made huge progress over the past decade, much remains to be done especially in the area of industry standards and open architectures. As CIOs struggle to integrate cloud-based services and applications from multiple vendors, standards and open source software are necessary to help cloud computing achieve the kind of market success enjoyed by the Web, e-mail and other widely used Internet-based services.