For the past couple of years, the trend in the IT world has been clear. The dominoes are falling, and the public cloud is taking over.
Still, there are some interesting nuances in the push toward greater public cloud adoption. Let’s take a look at where we see things headed, with the iconic year 2020 just ahead on the horizon. But first, a quick review of happenings in 2018.
2018 held a few surprises. The path to the future isn’t a straight line, of course. For instance—and some might not find this surprising—it’s noteworthy that serverless computing is still the rage. And despite predictions for the demise of private clouds, people are actually building more private clouds now than a year or so ago.
On the other hand, we thought there would be more multicloud implementations and that networking for multicloud environments would have become more straightforward. Unfortunately, however, it’s still quite complicated.
Now, on to what lies ahead.
Privacy concerns will remain a big thing. With people and elections being hacked, the implementation of GDPR, and questions of data sharing becoming more common, privacy has been at the top of everyone’s minds. We expect that governments will increasingly band together to establish common requirements for securing data and ensuring data privacy.
IT will be more involved in cloud governance. Initially, most growth in public cloud adoption within enterprises was driven at the grassroots level. But the rogue aspect of the cloud has pretty much played out.
Enterprises want a framework for their cloud implementations. IT will set certain boundaries, within which individual teams can choose best-of-breed solutions and do what they want with them. This light approach to governance will help IT exert more control over public cloud operations, and to embrace the cloud more tightly.
The cloud will become more complicated before it gets simplified. Cloud engineers, architects, and executives today control the bulk of IT budgets. That means that overall, enterprises’ technology environments will remain heterogeneous from a cloud and on-premises IT perspective—with the inevitable complexity that comes with heterogeneity. But we’re confident that longer term, the overwhelming advantages of simplification will motivate enterprises to seek solutions that make their cloud and IT operations easier to deploy and manage.
The skills shortage will become acute. It’s difficult to find skilled cloud architects who have run cloud operations on a large scale. That shortage will continue for at least the next 10 years. And when it comes to networking in the cloud, the skills shortage will be even more dire.
Cloud architectures will become more standardized. When the public cloud first appeared, the goal was just to get something, anything, to work. Solutions weren’t architected or designed for scale. They emerged, ad hoc, from the grassroots of enterprises.
As with the evolution of any space, the cloud world is beginning to establish standards. For instance, there’s great advantage to using a multicloud networking architecture that has been vetted across dozens of real-world enterprises’ environments, involving thousands of VPCs and connections with on-premises resources.
Enterprises are moving toward cloud patterns and architectures that scale and evolve. That way, they can avoid the expense and hassle of having to redo architectures that were designed simply to work in the moment.
Enterprises will continue to have choices of public cloud vendors. Amazon Web Services (AWS) is still way ahead of the pack, tackling big problems and innovating faster than anyone else. But the public cloud pie is growing rapidly, and both Microsoft Azure and Google Cloud Platform also continue to grow. Google and Azure are carving out specializations—machine learning/AI and desktop/business productivity, respectively—that make them best-of-breed options for various types of workloads.
Along with options, the existence of multiple public cloud vendors also means that multicloud deployments will become more prevalent, as well.
Increasingly, IT budgets will = cloud budgets. Already, Intel sells more CPUs to the top three public cloud vendors—AWS, Azure, and Google Cloud—than to combined CPUs sold to enterprises making private clouds. We see no reason for that trend to change.
Data-driven battles will come to a head. The future will be drenched in data (we’ll all soon learn what comes after petabytes). The data deluge will be driven in part by an explosion of data-generating connected edge devices.
The convenience offered by these connected edge devices will fight with privacy concerns. It will be interesting to see if tolerance for privacy tradeoffs—the technology equivalent of, “I’ll overlook lousy service at that restaurant as long as the food is good”—will diminish as companies and governments know more about each of us than ever. Or if the lure of ultraconvenience and ultrapersonalization will win the day.
Cloud will create greater similarities among disparate enterprises.The irony is that although greater public cloud adoption will enable enterprises to be more heterogeneous and more specialized, it will also push in the other direction, as well. For example, every company will be a software company. Every company will be a data-driven, cloud-enabled, AI-savvy company—in the same way that today, every company uses electricity, computers, and communications devices.
What will the cloud look like in 2020 and beyond? The harvest of the future is already taking root in the present. In other words, we expect pretty much the same as it looks now, only more and better.