Silicon Valley's next dynamic duo: Edge computing and cloud
- By 2022, more than half of enterprise-generated data will be processed or created outside of the cloud or data center, and edge computing will rise as one of the main topographies of data processing, according to Ed Anderson, VP and distinguished analyst at Gartner, speaking at Gartner Symposium in Orlando, Fla on Tuesday.
- Edge computing and cloud computing aren't competing architectures though: Edge computing needs the centralization and scale of cloud, and cloud needs the physical touchpoint of edge. Edge computing can solve latency, bandwidth, data location and autonomy issues associated with the cloud, he said. Latency is especially important and comes into play in systems such as autonomous vehicles, trading desks and mixed reality where millisecond lags can have costly effects.
- The challenge for businesses is building up the cloud and edge skills in their workforce, distributing security throughout edge systems, figuring out data ownership and moving from data collection and dissemination to intelligence driven by AI and machine learning, according to Anderson.
As edge computing becomes more ubiquitous in the enterprise, it will emerge as a classic duo, like peanut butter and jelly or Bonnie and Clyde, Anderson said.
The edge is the physical location where things and people connect to the networked digital world, according to Anderson. It can take many forms, including the connectivity hub fog computing. But what is common across edge environments is the fusion of the physical world to the digital.
The internet of things is a different, but complimentary part of the edge, a network of physically connected devices as opposed to the part where information is collected. While the dramatic growth in IoT devices has skyrocketed the amount of data being collected, not all that data is being used.
By 2021, there will be 21 zettabytes of data center traffic but 85 zettabytes of useable data; the area between this consumption and collection is where edge computing's opportunity lies, Anderson said. And the rise of edge computing is imminent. By 2020, more than half of enterprises will incorporate edge computing principles into their projects.
Edge computing can flip the centralized model of the cloud while using the cloud as a point of coordination or control. The geographic clustering of major cloud providers creates a lot of opportunity for the edge to extend the value proposition of the cloud into more remote environments, he said.
Businesses are already starting to build server capacity into low latency technologies, such as adding a server into the trunk of an autonomous vehicle or on a boat. By collocating the bodies generating data with the bodies consuming them, an organization can push information, such as a machine learning models, into points of data consumption.
Follow Alex Hickey on Twitter