As businesses continue to grow and expand, their need for data storage and processing increases as well. To meet this demand, many are turning to edge data centers. The demand for private and shared edge data centers has started growing and within the next 2-3 years, there will be a massive surge in the demand for edge data centers. In this post, we will discuss the types of edge data centers and what factors are fueling edge data center buildouts.
What is an edge data center?
Edge data centers bring IT capacity close to where data is generated or consumed. It reduces the latency to improve or add new IT services, minimizes network cost and/or bandwidth constraints, and alleviates data security and business continuity concerns.
The figure below shows the different types of data centers that can be deployed for edge computing.
Core data centers vs edge data centers
Core data centers are large or very large and provide services over large areas like several smaller countries or multiple metropolitan areas in a large country. Although they can provide sufficiently low latency for certain edge workloads, most core data centers are in remote locations simply too far from users to provide low-enough latency. The increase in high-speed connections and traffic volume can clog network access points and switches, slowing transmission.
Edge data centers carry out processing with both – low latency and reliability. They can be private or shared. Private edge data centers are owned by organizations that primarily do not sell cloud or IT services to other organizations. Shared edge data centers are owned by a service provider offering colocation or cloud services and provide services to multiple organizations.
The primary technical factors that determine whether data must be processed/stored at the edge or a more central location are latency, the volume of data, and criticality or business sensitivity. Requirements of the workloads are the ultimate factor defining the need for edge data centers. The following figure maps select edge workloads according to their typical requirements for criticality and latency.
The growing demand for edge data centers
A recent study by Uptime Institute reveals that in the next two to three years, the number of surveyed owners/operators using 20 or more edge data centers today will double from 9% respondents today to 20% in two to three years as shown in the figure below.
According to the study, there will be large deployments of hundreds or more edge data centers for many applications like telecom networks, internet of things (IoT) in oil and gas, retail, cloud gaming, video streaming, public transportation systems, and also due to the growth of large international industrial companies.
Currently, most suppliers of edge infrastructure are helping to build, supply or maintain 6 to 10 edge data centers per year. This is expected to grow to 21-50 annually in the next 2 to 3 years as shown in the figure.
Factors driving the demand for edge
The following factors influence the demand for edge computing infrastructure:
- The accelerating digitization across multiple areas and industries is expected to positively impact the demand for edge infrastructure.
- The launch of edge products by cloud companies extends their cloud platforms to distributed edge locations.
- Colocation data centers are likely to expand into more geographies in the future, including smaller cities and remote regions.
- Private 5G networks, in partnership with private edge data centers, are becoming a way to enable edge workloads, especially for industrial IoT.
- The growing popularity of augmented reality (AR) and virtual reality (VR) applications, like cloud gaming, demand low latency to improve user experience.
Image and source credits: Uptime Institute