With the launch of 5G, large amounts of data can be transferred at unprecedented speeds, opening up a whole range of new opportunities for innovation that have only been imagined as a possibility until now. The 5G ecosystem has not yet been fully realized and is projected to be fully realized in the near future, but visions of the IoT connecting machines and their learnings through everyday AI are still decades away from being realized. In order to meet the needs of their customers, a number of major wireless carriers are increasing their investments in fiber, small cells, and high-frequency spectrum, as evidenced by Verizon’s $45.5 billion success in the auction and AT&T’s $23.4 billion success in the auction.
Having said that, the network and its pipes are just one small part of the IoT universe as a whole – if we take a look at what turns data into gold, it is the uber-fast computing power of the cloud. Creating edge data centers as well as shifting computing to the point of use is when 5G will truly come to life and we will finally be able to deploy IoT at its full potential. The question then becomes, how can data centers evolve in order to accommodate the next generation of users? Even though clients are preparing for 5G technologies and devices, the industry is still working on a definition of what exactly an edge data center is. This is because clients are preparing for 5G technologies and devices.
As data centers evolve, they will be less defined by size and more by proximity to end users and their nimbleness to process, move, and store data.
Machine-to-machine applications and real-time data analytics based on SDNs and network virtualization are the foundation for 5G, which is powered by high-density servers located in today’s data centers. This network of increased computing power has led to hyper-scale data centers and cloud facilities. Those with legacy data centers must (if they haven’t already) migrate to cloud providers or upgrade infrastructure to support future 5G innovations.
As data and network, demands increase, 5G’s impact on data centers will remain centered around optimizing efficiency, reliability, resiliency, and security.
In addition to consuming more power, generating more heat, and requiring new approaches to how equipment is configured and cooled, next-generation solutions also present conditioning challenges. The cooling system bears the brunt of greater data processing.
Energy-intensive systems have been experimented with to keep cool. High-density computing now uses liquid immersion cooling and direct-to-chip cooling. Water usage effectiveness (WUE) will be given more attention, which is already a hot topic. Chips that need less cooling are an alternative.
Cooling and power go hand in hand. They work together to maintain power usage effectiveness (PUE). A high-power-consuming industry will see even more power demands from 5G and IoT. To manage the expected rise in computing power, hyper-scale data centers are already focusing on decarbonization. The green data center market will grow from $49.2 billion to $140.2 billion by 2026. Microgrids, battery storage systems, and hydrogen fuel cells are other initiatives beyond renewable energy credits. Nuclear power, which has a proven track record for innovation hubs, research facilities, and the like, is also gaining interest as a mature, clean technology. A data center near a nuclear power plant or advanced small modular reactors (SMRs) may be used to increase speed to market.
Increasingly, data centers will be defined by their proximity to end users and their nimble ability to process, move, and store data. In 2025, Gartner predicts that 75% of enterprise-generated data will be created and processed outside of traditional data centers. Providing latency, resiliency, and accessibility to consumers requires a network philosophy and architecture that support storing, processing, and converging near end users.
Data management will be boosted by the proliferation of 5G. Edge will emerge as data gravity becomes stronger, making it harder and more costly to move.
Computing is evolving at the edge faster than the infrastructure can handle, and its proliferation requires smart responses immediately. Edge data management, security, and curation must grow significantly in scale and sophistication to realize machine learning.
Infrastructural hubs make use of existing resources. A natural location for edge data centers can be a geographically interconnected telecommunication tower with access to power and fiber. The companies operating tower sites, such as American Tower and SBA, are actively hosting data centers. Together with Dell and FedEx, Switch, Dell, and FedEx deployed edge data centers and cloud solutions throughout FedEx distribution centers at the end of 2020. With 5G, infrastructure applications will likely be closer to renewable power sources and fiber to support advanced infrastructures, such as autonomous vehicles and industrial IoT.
There are resources available to make IoT a reality. It is unlikely that one edge data center will be enough to accomplish the job alone. In order to accelerate deployment, owners/ operators must think strategically about designing and building edge data centers that can leverage scale opportunities and speed to market at the same time. Each edge data center will be a unique and customized solution tailored to the organization’s needs and the needs of its location. The success of edge deployments depends on flexible, scalable solutions. Among the most significant factors to consider are:
The design of an enterprise’s edge facilities is heavily influenced by latency requirements and the ways in which data will be consumed, stored, and transported. For real-time data analytics and response, will the data be processed and stored in the enterprise or in the cloud? Data centers are designed, located, and sized according to these factors.
Data requirements will continue to evolve as 5G evolves. To accommodate the scalability of data growth will require, edge data centers will require modular facilities. Currently, supply chain crunches only make this need more pressing.
Coordinating a complex network of interdependencies is key to ensuring edge infrastructure deployments are completed on time and within budget. To design, permit, zoning, prefabricate, ship, install, and commission a site, a team of partners and vendors is needed.
Maintaining a consistent customer experience across all locations will require visibility into the entire network of edge facilities deployed and maintained by multisite enterprises. Monitoring and diagnostics of edge operations will be a key focus area for edge owners/operators.
From site identification and due diligence to permitting, engineering and design, installation and commissioning, consider the entire edge data center deployment process. Owners and operators may wish to hire a project manager and a partner to manage the deployment and mobilize at the required speed and scale.
The extreme dynamics of 5G will continue to be complex for some time to come. Managing this fluid, interconnected web of dependencies won’t go away. Since obsolescence cannot be prevented, everyone involved will need to make a decision about the standard technology philosophy and architecture that will allow for growth and agility. There will always be something new to learn, always something new to pursue, so we need to accept that we will always be behind the technology curve, always trying to stay ahead of the curve.