Edge computing on solid footing — understanding the three modes of edge deployment
By Tony Grayson, General Manager, Compass Quantum
There are many observers in the data center industry that have closely followed the progress and promise of edge data centers. Could the edge deliver on its potential to solve our computing, storage and high-performance compute needs? In most articles, the edge is touted as the cure-all, able to meet all requirements for data storage and computation, with the benefits of lower latency and increased security.
On the surface, it makes sense. What company doesn’t want low latency, lower network costs, and greater server security inside their fence line? But in reality, the edge has not yet materialized as a widely adapted data storage solution. Scour the internet or read any number of blogs and you will find multiple explanations for the delay, such as the lack of a transport network, the failure of 5G to reach initial expectations, and the absence of a killer application. All are true to some extent.
But we are overcoming these obstacles, and there is a sunny forecast for our industry. After years of waiting, the edge revolution is finally here as a result of the focus on digital transformation bought about by the pandemic. This evolution has been called a few different things — the next industrial revolution, Web 4.0, or Industry 4.0. Whatever you call it, it’s reached an inflection point that indicates it’s here to stay.
As a result of this change in focus, companies are making a concerted effort to figure out how to use machine learning, artificial intelligence, and big data to bolster global production, support remote operations, optimize manufacturing and support consumer-focused applications. The global Industrial Internet of Things market is expected to realize a compound annual growth rate (CAGR) of 23.2% between 2023 and 2030, up from $394 billion in 2023.
Defining the edge and the architecture to support it has been a lively discussion topic at conferences and events for years. But the conversation has never felt real because there was no real use case. Now, with a real and well-understood need, edge computing is starting to take shape with three distinct models:
- Hub-Hub-Spoke Model. In this model, fiber provides the main form of transport. The first hub is a data center that offers an onramp and cloud peering point that connects to a second smaller hub. The second hub serves as an aggregation point closer to the customer or endpoint, creating economies of scale at 1 to 5MW. The second hub then spokes out to smaller deployment locations spread across rural and metropolitan areas to reduce latency and network transport costs for the platforms under development in the 100 to 500kW range. This is also called Near Edge and Far Edge.
- The Fixed Wireless Model. This model is based on the premise that for many locations, fiber can be too expensive and take too long to deploy. So, companies should make their platforms based around fixed wireless transport from the endpoint to a base station that has local compute. This smaller aggregation point (100 to 500kw) is backhauled to a location for peering and onramp as with the first use case. Today this wireless could be 5G and 5G private networks, but would follow with 6G, etc.
- The Death of the AV Closet Model. This model shows the promise of being the most universally deployed. In the past, offices, hospitals, and other enterprises relied on smaller onsite deployments inside their facility for all of their rack needs. However, as modern computing needs become more power-hungry, requiring more cooling, and the number of racks continues to increase to support more modern platforms, space requirements become an issue due to a need for on-premise or near-premise locations for additional racks. Moreover, there is an increasing need for AI/ML in hospital and industrial applications. While these locations might be in retail data centers now, customers are looking for other options where they have more control.
As a result of the new landscape and our need to pursue new storage solutions, edge is finally starting to deliver on its promise to deliver low latency, security controls, and cost-effective computing and storage. These benefits are not reserved for large corporations alone but for smaller and medium-sized enterprises.
Businesses of all sizes can take advantage of reduced network costs and improved data security. This infrastructure can transform the healthcare, finance, retail, and transportation sectors by enabling real-time decision-making, predictive maintenance, and improved customer experience. edge will also be critical in enabling smart cities, smart roads, and connected communities.
The edge computing revolution is no longer just a concept but a reality that organizations worldwide should pursue. With the growth of IoT, AI, and big data, the edge is poised to become a crucial component of the digital landscape, providing fast, secure, and cost-effective computing solutions. The hyperscale data center or cloud is not going away, but edge computing provides an alternative worth exploring.
About the author
Tony Grayson leads Compass Quantum, a division of Compass Datacenters dedicated to delivering turnkey, modular data centers and giving customers the flexibility to remotely monitor, manage, and operate these locations. Before joining Compass, Tony was an SVP at Oracle, where he was responsible for their physical infrastructure and cloud regions. He has also held senior positions with AWS and Facebook. Before embarking on his data center career, Tony served for 20 years in the United States Navy.
DISCLAIMER: Guest posts are submitted content. The views expressed in this post are that of the author, and don’t necessarily reflect the views of Edge Industry Review (EdgeIR.com).
IBM unites with Juniper Networks, Nokia, Spirent to power, automate next-gen networks
cloud | Compass Datacenters | Compass Quantum | data storage | edge data center | interconnection | networks