Where to Put IOT Data and Analytics – In the Fog or the Cloud?


With 50 billion connected devices estimated in the IoT by 2020, its impact will be felt within every facet of enterprise IT infrastructure. A recent Forbes article explores how the IoT will shake up the future of the data center, particularly by making the edge of the corporate network a lot smarter.

With so many connected devices and huge amounts of data being generated, it’s not realistic to expect that the corporate data center will be the hub of the data storage and analytics strategies for digital business. Everyone knows how hard it is to backhaul data to a centralized data center to try to gain timely business and operational insights from centralized analytics applications. By the time you get the answers back from the data you’ve analyzed, there is new information that outdates what you’ve just learned.

Securely pushing data and analytics to the edge of the corporate network, closer to the users and applications that need real-time access to them, solves this issue. This is exactly what “Fog” computing promotes. Fog computing enables the analysis and management of IoT data locally, at the edge of the network. The premise is that by putting the compute and analytics capabilities closer to the edge, where much of the data is created and accessed, fog computing enables you to decrease the distance the data must travel over the network. As a result, you reduce latency and bandwidth costs and derive immediate insights for more timely decisions.

But then there is the “camp” that assumes the cloud will be picking up most of the analytics traffic. IDC in its recent report, “IDC FutureScape: Worldwide Cloud 2016 Predictions,” estimates that the cloud will be the preferred delivery mechanism for analytics and will enable public information consumption to increase by 150%.

At Equinix, we believe that a distributed, edge-based IoT infrastructure can benefit from both fog and cloud computing paradigms, especially if they are empowered by an Interconnection Oriented Architecture™ (IOA™). Equinix’s IOA™ shifts the fundamental IT delivery architecture from siloed and centralized to interconnected and distributed in support of edge and cloud computing infrastructures. And the Equinix Interconnection Platform™ provides the critical building blocks to implement this architecture – a global footprint, dense cloud and service provider ecosystems, and the ability to integrate data and analytics at the edge.

The Forbes article rightly points out that the vast majority of industrial and consumer endpoints are not yet connected, which is why enterprises need to think of interconnection first, prior to deploying either a fog- or cloud-based IoT strategy. And, along with interconnection requirements, the article recommends that you look at the types of data that are being analyzed, the amount of data that needs to be managed and the cost of pulling it all together. As a matter of fact, a recent Forrester study shows that optimizing your interconnection strategies can help you reduce costs and gain a 300% return on your investment.

The other thing to consider is that an IOA can help you deploy a flexible and agile data and analytics infrastructure that will enable you to leverage both the edge and the cloud. And the Equinix Interconnection Platform makes it possible to mix and match fog and cloud computing where it makes sense ̶ by where specific types of data and analytics reside, based on time-sensitivity (mobile healthcare apps), the mission-critical nature of the information (security), or if you are gathering data from just one or multiple sources (supply chain or customer behavior analyses).

Learn more about how Equinix’s Interconnection Oriented Architecture is transforming today’s data centers.