Thank you for Subscribing to CIO Applications Weekly Brief
Edge Computing - Evolution or Revolution?
Rob Nash-Boulden, Director, Data Centers, Black & Veatch
By definition Edge is "distribution of compute or storage to the periphery". As a result, Edge is sometimes described as a form of Distributed Cloud. Cloud will enable Edge In the same way that virtualization, software, and networks enable Cloud. There is consensus that Edge is difficult to define – much in the same way that many struggled to initially define Cloud. Experience tells us that technologies evolve, shift and adapt – rather than switching. Thus, the trend toward Edge computing will be based on the limitations of Cloud and much more of an Evolution than a Revolution.
Fast paced growth in IoT, machine learning, and big data are a catalyst for technological advances in processing, networking and analytics. This combination is creating and enhancing opportunities for revenue generation, new services, and improved customer experience. Edge will test the limits of latency and bandwidth causing current best effort telecommunications networks, Cloud and data storage capabilities to best stretched beyond their limit. Increasingly, analytics and processing will be performed as close to the source of the data as possible.
The evolution of Edge networks will be closely tied to the introduction of new services and solutions in device and data Security. If Edge is the cure for latency, then Security is the cure for the privacy requirements and regulation that distributed Edge computing networks will require. Zero trust policies, authentication before connection, and secure software defined perimeters will be the death of VPN’s and will pave the way for IoT devices to be managed in the way they were intended rather than as gateways into to secure networks. Blockchain technologies and their associated traditional, digital, and crypto currency families will provide the ability to exchange goods and services for “money” for billions in the world without access to banks.
This seems like a good a point as any to introduce an important factor in the evolution to Edge - Data Gravity. Software and Applications are attracted to each other. The more data there is - the greater the attractive force pulling applications and services to associate with that data. This attraction is partly physical in the sense of network characteristics like latency and throughput. Data gravity and mutual attraction can also be thought of in terms of network effects more broadly. Large and rich collections of information – whether images, data about images, video, purchase data, or just about anything – tends to pull in more and more services that make use of that data. The opportunity that data gravity affords IS HUGE and allows large Cloud and Edge providers to extend their dominance. As users store more data, data gravity makes it easier and more efficient for users to use additional services to leverage that data more fully. This creates a “virtuous cycle” where data gravity creates opportunities for more services to be used, which generates more data that then creates more services to be consumed.
At least one pressing question remains – when will we see a measurable shift to Edge computing? If we consider the continued frenzied pace of hybrid cloud adoption and digital transformation, what could possibly cause use cases to evolve that will in effect be the tipping point? I subscribe to the theory that it will be a combination of at least three things:
1. When there is a fundamental shift away from merely consuming data toward meaningful uses of this data via analytics. This will include more balance between requirements for data “uploads” from devices, rather than the current reliance on “download” crippling the current best-effort network. Project this further and consider the compute, storage and network requirements for machine learning, augmented reality, and artificial intelligence and it becomes apparent that things will need to change – Cloud and Edge are going to be familiar counterparts within networks.
2. Highly latency sensitive AND bandwidth intensive use cases are developed. Transportation is one likely possibility. Healthcare is another. Some of these applications are not terribly sensitive to latency and bandwidth today, but it is relatively easy to imagine that changing from a convenience to a matter of life and death.
3. Follow the money. Think about use cases that today are not terribly latency sensitive such as fantasy sports for instance. Layer on the development of this use case from leagues players compete for fun to those where money is wagered. Layer that further and project the impact and financial opportunities that may exist if real-time betting on live sports were to come to pass. Now, an application that wasn’t particularly sensitive to latency becomes incredibly sensitive. Layer in the fees that players and the application providers are potentially willing to pay or invest to enhance that experience and you have a “perfect storm” for investment into Edge.
In conclusion, Edge will not be the end of Cloud. Rather it will be another important piece of smart integrated infrastructure. 5G networks will be enablers for Edge helping to solve for apportionment of bandwidth and latency. Security, priority, data curation, data gravity, and sensitivity to latency are all key trends. Use cases and financial investment will work hand-in-hand to enable distributed networks to morph into the shape of the future both in the Cloud and at the Edge – get ready for the Evolution!