The Changing Definition of Cloud Computing

Once upon a time, it was pretty clear what people meant when they talked about “the cloud.” They were referring to servers and/or data storage that ran in the cloud. Today, however, the cloud could mean many other things, as well. Appreciating how the definition of cloud computing and cloud services has evolved over the past decade is critical for understanding how organizations can leverage the cloud today--as well as for predicting the future of cloud computing.

The Original Cloud: IaaS, SaaS and PaaS

When marketers and IT managers started talking excitedly about “the cloud” in the mid-2000s, they were referring to three types of services: IaaS, PaaS or SaaS. Actually, early on, they were referring above all to IaaS, which at the time involved virtual machines and storage that you connected to via the internet. That was the main service provided by the first big public cloud companies, like Amazon Web Services. When AWS launched in 2006, its first major services were EC2 virtual machines and S3 storage.

PaaS and SaaS were models that had largely predated the advent of IaaS. However, in the early days of PaaS and SaaS, no one called them “the cloud.” When Salesforce launched in 2000, its tagline was “no software.” It didn’t mention the cloud. Likewise, PaaS tools like Heroku were usually thought of as stand-alone developer tools that happened to feature hosting services.

Still, by 2010 or so, folks had begun referring to SaaS and PaaS as “the cloud,” as well. The definition of cloud computing began to include all three of these things, especially after the major public cloud providers started integrating SaaS and PaaS solutions into their portfolios.

The Evolving Menu of Cloud Services

Since then, cloud providers have grown into conglomerates offering much more than standard IaaS, SaaS and PaaS.

Consider, for example, just a few of the disparate types of services now offered by AWS:

That’s just a partial list of the services offered by AWS today. The full list is much longer, as are the lists for Azure and Google Cloud services.

All of these services follow an SaaS model, and some (like Managed Blockchain) are essentially a form of IaaS. RoboMaker is also more or less a PaaS.

What’s remarkable about these types of services, however, is they go far beyond anything that most folks were envisioning from cloud providers a decade ago. Back then, being able to host an application on infrastructure managed by someone else was a big deal. Today, we’ve become accustomed to being able to manage satellites from the same cloud provider that stores our data backups and hosts our applications.

The Evolution of IaaS

It’s worth noting, too, that IaaS has changed significantly over the years. Gone are the days when basic virtual servers and storage were the be-all, end-all of cloud-based infrastructure.

Today, you can use the cloud to rent a physical server, circumventing the limitations of having to rely on a virtual environment to host your cloud applications. You can also host containerized applications running on Docker, Kubernetes or (in some cases) another container framework of your choosing.

These were solutions that were basically unimaginable circa 2010, when Docker did not yet exist.

The Future of the Cloud: Performance and Cost Cptimization

Given the many ways in which cloud services have evolved and become broader over the past decade, it’s worth thinking about where we go from here.

In some ways, it’s hard to imagine what else cloud computing providers can add to their portfolios that is not already there. From data analytics to robotics to IoT, they have most niches covered. Although certain types of new SaaS services are bound to keep cropping up, I don’t think the future of the cloud lies primarily in offering more and more services.

Instead, I suspect that performance and cost optimization will be the next big frontier for cloud computing. Cloud performance faces many of the same bottlenecks today that it did a year ago. In most cases, your cloud still performs only as well as your data connection.

But there are signs that cloud providers are moving to tackle this problem via solutions like faster network interfaces and edge computing solutions.

At the same time, more and more vendors are rolling out tools to help users optimize the costs they pay in the cloud. While the cloud providers themselves have long offered cost-analysis and prediction tools, third-party solutions like Cloudability and Nutanix have entered the fray, promising easier-to-interpret cloud cost analyses. (And, let’s face it: The cloud providers don’t have much of an incentive to build tools that really help you spend less on your cloud bill.)

So, while trendy new SaaS and IaaS offerings from cloud providers are likely to continue to make some headlines for years to come, I suspect the real innovation (and competition) in the cloud will be about performance and cost optimization. Most organizations are already using the cloud, and they’re looking for ways to use it more efficiently and cheaply. Cool new services will be less important.