Deployment models at the edge

Uptime Institute research suggests that a majority (60%) of workloads deployed at edge facilities are hybrid applications that rely on a centralized location for back-end processing and storage. Cloud providers are complementing their centralized architectures with capabilities that enable customers to deploy public cloud software in many locations. On the facility side, demand for small-scale edge data centers (tens to hundreds of kilowatts) is growing but has not met the initially high expectations. Conversely, larger megawatt-scale data center buildouts into new geographical (edge) regions continue at a rapid pace. This report discusses the prevailing deployment models for data centers facilities and IT at the edge, together with key considerations.

KEY POINTS

• Organizations commonly seek to deploy an asset light, consumption-based architecture for their edge workloads (covering both IT and facilities). They have an increasing range of options from both cloud providers, at certain colocation facilities, and on-premises.
• The high expectations relating to site-level resiliency are holding back the deployment of smaller far edge data centers due to operational and capital cost.
• Most current workloads can tolerate the added network latency from metropolitan and regional facilities. Far edge facility investment has remained lower than many expected because of this.
• Some organizations with “cloud first” strategies will still need to manage large numbers of small on-premises data centers, although these may be limited to a few racks of IT.
• Upcoming regulatory performance thresholds may cause some older edge data centers to be scaled down (or, in extreme cases, closed down) with workloads migrated to more efficient colocation or cloud facilities.