Get Ready for Distributed Cloud at the Edge

As the amount of data soars, interconnecting compute and storage at the edge is essential for high performance and low latency

Herbert J. Preuss

Internet of things (IoT) and artificial intelligence (AI) use cases are already generating massive amounts of data “at the edge” – on mobile and IoT devices, in buildings, on farms, etc. According to recent research by Frost & Sullivan, by 2022, 90% of industrial enterprises will be using edge computing, with the .[i] And as 5G continues to roll out, we will see so much data generated at the edge that transporting it to centralized processing facilities will no longer make sense from a network performance or cost perspective.

5G will also become the foundation for smart cities, autonomous vehicles, major advances in manufacturing and logistics, and so much more. This will lead to an exponential growth in machine-to-machine (M2M) communication – according to Cisco, by 2023, there will be 14.7 billion M2M connections, comprising half of the global connected devices and connections.[ii] The importance of this is that while consumer applications can tolerate latency well above 20 millisecond (ms), most M2M applications require sub 20 ms latency. To ensure this, compute and data storage must be pushed closer than ever to users. We are seeing two closely related trends that are coming together to solve this challenge: edge computing and distributed cloud

90%

By 2022, 90% of industrial enterprises will be using edge computing, with the multi-access edge computing market set to reach $7.23 billion by 2024

Gartner - Top 10 Strategic Technology Trends for 2020 Edge

New business requirements and technical capabilities are driving more computing and communications power toward the edge of the network”, according to this Gartner Research report. "Edge architectures and technologies will be an essential component of innovative products and services.

Read More
Callout pic

Edge computing

Empowered Edge was named by Gartner in their “Top 10 Strategic Technology Trends for 2020.” Edge computing is a topology where information processing and content collection and delivery are placed closer to the sources of the information, with the idea that keeping traffic local and distributed will reduce latency.[iii]

In the past, any data collected by edge devices, such as a smartphone or manufacturing equipment, had to be “backhauled” to a centralized processing location. In edge computing, compute and storage are distributed closer to where the data is generated and consumed, often by adding infrastructure via proximate servers, routers or datacenters. In addition to the use cases mentioned above, there are many current and future use cases for edge computing, such as connected robots and  autonomous things, smart manufacturing and intelligent home systems. With 5G as the essential enabler, edge computing will put the “smart” into many more technologies.

Distributed cloud

Distributed Cloud was also named by Gartner in their “Top 10 Strategic Technology Trends for 2020.” Distributed cloud refers to the distribution of public cloud services to locations outside the cloud provider’s physical data centers, but which are still controlled by the provider.iii

Essentially, distributed cloud applies the idea of edge computing to public cloud infrastructure. Consider the challenge that cloud service providers (CSPs) have in meeting the needs of latency-sensitive use cases. If a CSP has a cloud availability zone in London, businesses in many areas of Europe may not be able to access that CSP’s services with sub-20 ms latency. Further, a business that is employing edge computing strategies wants to be able to manage the distributed resources using a simple, centralized, cloud-like strategy. One of the early examples of distributed cloud to cloudify the edge is Amazon Web Services’ Outpost.

Most machine-to-machine (M2M) communications require sub 20 ms latency. Edge computing and distributed cloud can help solve this challenge by pushing compute and data storage closer to users

How 5G impacts these trends

To solve these challenges, CSPs are following the edge computing model and partnering with telcos and network service providers (NSPs) to distribute localized compute and storage capacity around the world. And 5G is only accelerating the convergence of these edge computing and distributed cloud trends. For example, AWS is partnering with Verizon, Google is partnering with AT&T and Microsoft is partnering with Vodafone on edge computing/distributed cloud solutions to name a few.[iv]

While still in the early stages of deployment, the potential benefits of these edge computing/distributed cloud solutions are vast. They will enable the management of distributed applications and data across heterogeneous infrastructure while delivering comprehensive and integrated compute, storage, networking and security – all centrally and seamlessly managed.

Solving the hybrid multicloud challenge

Even as edge computing and distributed clouds mature, one important hurdle remains. Businesses often wish to rely on multiple clouds in order to enable increased innovation, avoid vendor lock-in, and allow the use of the best cloud for each workload. Creating a multicloud environment requires businesses to deploy virtual infrastructure in each cloud, and each infrastructure needs to be managed separately. And, as today’s public clouds are designed, applications deployed in multiple distributed cloud infrastructures cannot be easily connected. In fact, this challenge has led many senior IT decision makers to feel their infrastructures are getting more complex today, not simpler. According to recent research by Enterprise Strategy Group, digitally mature organizations were 3x more likely than those with no digital transformation initiatives to say IT is significantly more complex today.

Distributed cloud at the edge will help pave the way to true multicloud in the future by enabling enterprises to seamlessly manage on-premise services and services spread across different clouds as one logical cloud with minimal infrastructure duplication.

Vendor-neutral interconnection solutions such as those found on Platform Equinix® are also helping to solve the hybrid multicloud challenge. Businesses can connect to most major clouds in strategic metros within 1-2 ms latency, simplifying IT management via a single globally consistent platform. Our 210+ International Business Exchange™ (IBX®) data centers are interconnected physically and virtually across 55+ global metros in 27 countries on 5 continents. Proximity to the edge in most metros is less than 10 ms latency from end devices. Equinix Cloud Exchange Fabric® (ECX Fabric®) on Platform Equinix interconnects dense ecosystems of clouds, networks and businesses over high-speed, low-latency, virtualized connections around the globe. This enables CSPs to distribute and scale compute storage wherever they need to be, while enabling enterprises to access multiple, interconnected CSP services.

For more information on the power of computing at the edge, read Gartner – Top 10 Strategic Technology Trends for 2020 Edge

 

You also may want to read:

 

 

[i] Frost & Sullivan, 5G and Edge Computing—Cloud Workloads Shifting to the Edge, Forecast to 2024, Aug 2020.

[ii] Cisco, Cisco Annual Internet Report (2018–2023), 2020.

[iii] Gartner Ebook, Top 10 Strategic Technology Trends for 2020, Edited by David W. Cearley, 21 Oct 2019.

[iv] STL Partners, AWS, Azure & Google at the edge: How much of it is telco edge computing?

Cloud providers are following the edge computing model and partnering with telcos to distribute localized compute and storage capacity around the world