DIGITAL SOLUTIONS

Are we entering a world of Edge computing?

Larry Socher, global managing director for infrastructure services at Accenture

May 27, 2019

Larry Socher, Global Managing Director for Infrastructure Services at Accenture

2019: it’s not just about cloud

Despite what all the hype about cloud may lead you to believe, it’s about enabling business outcomes, not necessarily the concept of cloud computing. When we speak to clients, they only care about how cloud can enable their applications, which in turn, enable their business.

Put simply, if cloud-enabled apps make their business more agile, then they’re all for it. Similarly, they do not focus on infrastructure as an ends in itself, but how they can use agile and software defined to transform infrastructure from a “bottleneck” to “enabler” of the business, and that’s what they get excited about.

Application led journey to cloud

Most cloud and infrastructure conversations we have with CIOs start with ‘what is your application strategy, where are your apps, and how do they work’. As established businesses such as retailers or banks embark on digital transformations to better compete in today’s market, they need to tap into cloud and transform their infrastructure in order to support the application modernization required to enable the change.  As we like to say, their journey to cloud and new IT is “application led, cloud and infrastructure enabled, secure, powered by intelligent operations.”

Data is on the move

As our clients adopt cloud native development and modernize and integrate their applications through micro-services, one of the key questions they must ask themselves is where does data reside and what applications need to access it. Regulatory changes such as GDPR and the growth in public understanding of the power of data have businesses reconsidering where they generate that information, where they store it, and why those two are, currently, different. Do they need to be?

As customers talk about wanting to be more agile and dynamic, part of their transformation is going to be about where and how they analyse and draw insights from their data. Traditionally data has been aggregated into large, centralized data warehouses for analysis.  Over time these warehouses have been migrated into the cloud with innovations such as AWS RedShift.   However, being agile increasingly means being able to react quickly and effectively.  As we hit an inflection point with the Internet of Things (IoT) where we shift from proof of concepts to scaled deployments, we are about to see a shift away from data existing in centralized repositories (data gravity), to a world where the majority of data resides at the edge.  As network price performance does not keep pace with compute and Moore’s Law, it will not be feasible to move all of the exponentially growing edge data into centralized clouds to process.  Instead, we must start to process more data in motion at the edge of our environments, being analysed and actioned in branches, devices and other endpoints, closer to where it is created. We must also minimize the amount of data being moved to central repositories for batch analytics, artificial intelligence, and machine learning analysis.

The ability to drive analytics, AI, and ML and process data in motion at the edge of our networks to support IoT will require significantly more compute power at the connected mine, retail store, or digital bank.  As multiple IoT applications will need to coexist and share compute, storage, and network resources, for example predictive maintenance video analytics and autonomous vehicles for a connected mine, we anticipate the increased use of containers orchestrated by Kubernetes to support IoT applications.  As distributed compute becomes more critical for operating in the digital business, not only must these distributed platforms be managed and maintained, but must be secured to protect the business.

2019: the year of secure business at the Edge

It’s a cliché to talk about constant change being the only guarantee, but we’re operating in a continuous transformation era – everything is changing. As need to generate, analyse and action data at the Edge increases, new solutions will be required for enterprises to process, secure, and transport their data.  Accenture is partnering with VMware to develop Managed Edge solutions providing integrated compute, storage, and networking up through the container to support multiple IoT middleware and applications that can process data at the edge.

While security has often been an afterthought, given the sensitivity of the data, we have been very focused securing the applications, data, compute, and networks from the ground up, providing end-to-end security from IoT devices through cloud with an integrated DevSecOps model.  Just like VMware, we believe security has to be embedded in from day one.  We are working closely with VMware leveraging innovative technologies like NSX with VeloCloud to automate and micro-segment device to cloud networking, providing integrated policies for security and QoS to support IoT applications.  This allows us to extend micro-segmentation to the Edge, so that the endpoints can be secured where they exist, whether they are a mobile device used by a retail assistant to check stock levels, an autonomous vehicle at a connected mine, or a wind turbine miles in the North Sea.

These integrated Managed Edge solutions allow our clients to analyse and action based on data, without losing time having to collate it all in a centralised point, while remaining completely secure. While this is a concept that has been discussed for some time – we are leveraging our partnership with VMware to make it a reality.

It’s less about where it gets done, just that it’s done

Ultimately, what matters is that the enterprise, whether a retailer, a bank, a hospital, or whoever it might be, can meet its business objectives. Today, that means having the cloud and infrastructure to enable the application to quickly and cost-effectively enable the digital agenda of the business. Tomorrow, that means securely extending compute, storage, and networking to the edge to support the new processing requirements of the Internet of Things.

Watch video

In the same category