How Will 5G Impact Traditional Data Centers?

The hype about 5G, the next generation of wireless networking, continues to accelerate, with an endless flow of announcements from carriers, communication service providers, equipment manufacturers and standards bodies. There’s no question 5G holds tremendous promise, delivering major advances in data transfer speeds, latency, connectivity, capacity, reliability and mobility.

Beyond that, questions abound.

For instance, when will carriers be able to offer organizations 5G capabilities, coverage and services that 4G LTE can’t deliver? To prepare for 5G, they’ve been furiously working to shift to software-defined infrastructure, and they consider B2B markets critical to recouping these investments and creating new revenue streams. On the buy side, when will use cases be sufficiently compelling to convince businesses to upgrade to 5G, particularly when they’re still getting good traction with 4G LTE?

In its 2017 “Industry Impact of 5G” survey, Ericsson asked C-suite executives to list key barriers to 5G adoption. Sixty-two percent said it was too soon to know 5G’s real benefits. Other noteworthy hurdles included the lack of standards (cited by 76 percent) and use cases (47 percent).

At this point, there are a lot of “ifs” with 5G, says Mike Fratto, senior analyst at 451 Research. “Carriers are just now starting to run trials, while providers have deployed some 5G radio frequency equipment, based on cell locations,” he says. “But carriers are still trying figure out how to run a 5G wireless network.”

In June 2018, the 3rd Generation Partnership Project, or 3GPP, approved the first standard specs for standalone 5G NR, on the heels of 2017’s nonstandalone standard release. Initially, 5G will lean on 4G LTE, which IT staff has used effectively to gain some of the functionality that 5G is designed to deliver.

5G Use Cases on the Horizon

“There’s no doubt that much of the 5G activity has been focused on investments from service providers and equipment manufacturers,” says Nick Lippis, co-founder and co-chairman of the Open Networking User Group. “However, more IT leaders are starting to make plans for 5G, which include determining its impact on their data center architecture, procurement strategies and the solutions they’ll roll out.”

Industry observers expect 5G’s speed, capacity and latency attributes, and services such as network slicing — which allots users a virtual piece of the network with personalized properties — to eventually play big in Internet of Things communications, artificial intelligence services and advanced analytics.

“The first thing IT sees is the ability to accelerate the elimination of a lot of server, storage and network hardware from their data centers,” says Lippis.

The combination provides a flexible path to Infrastructure as a Service, while freeing IT departments from the resource-intensive job of managing hardware-centric infrastructure. Instead, Lippis says, IT ops teams can focus on work that impacts business outcomes, from integrating solutions, strengthening security and creating new business models to automation and orchestration projects.

Already substantial, the number of organizations using cloud-native services and cloud-enabling technologies continues to grow. Public, private and hybrid cloud models and the wealth of cloud services give IT teams the flexibility to run applications and workloads where they make the most sense. 5G’s contribution, meanwhile, is to eliminate the latency issues that plague WANs, speeding wireless connectivity to cloud-based assets, as well as data transfer and download rates.

IT Spending Is Shifting to the Cloud

According to a 2018 Gartner report, 28 percent of investments in four key enterprise IT segments — system infrastructure, infrastructure software, applications software and business process outsourcing — will shift to the cloud by the end of 2022.

Not surprisingly, application software will continue to make up the biggest chunk of cloud shift through 2022, but system infrastructure will be the most rapid cloud-shift segment over the forecast period, jumping from 11 percent in 2018 to 22 percent in 2022.

Earlier investments in data center hardware, server operating system software and virtualization can impede more immediate moves, but as contracts come up for renewal, more IT ops teams will make the leap to the cloud. Gartner estimates that, by the end of 2020, IaaS investments will account for 39 percent of total data center system spending.

5G Will Compel Data Management Strategies

Given 5G’s potential role in fueling IoT, AI and advanced analytics initiatives, organizations that upgrade will also need an upgraded data management strategy. Businesses will need to be able to pull actionable, on-the-fly insights from the voluminous data they’ll generate, not only to justify investments, but to advance business objectives.

As early as the end of 2019, the IoT will be cranking out upwards of 500 zettabytes of data annually. To conduct real-time analysis, organizations need to reverse the analytics workflow and take compute and analysis to the data rather than waiting for it to come home. 

Enter edge computing, a key element of a distributed data center architecture that exploits 5G bandwidth and new service stations located between on-premises and cloud computing resources. In this model, endpoints or nodes sit at the very edge of the network, collecting data transmitted from local IoT devices and sensors. They then process or analyze it on the spot, save it and push it up to the data center.

Organizations can also purchase edge computing as a cloud service. Consider, says Fratto, an oil and gas company that has IoT sensors on all its rigs. They’ll typically send their data to the cloud for processing rather than to their own data centers.

Edge endpoints range from highly intelligent devices to nodes that wake up only when called upon to perform simple tasks. Taking a page from branch models, which today typically rely on a whitebox server or similar device for computing needs, providers are working to consolidate functions for edge computing. For example, says Fratto, they might consolidate software-defined WAN, firewall services, Active Directory and Private Branch Exchange functions on a single appliance, or run two appliances in high-availability mode.

By 2022, according to a 2018 IDC report, more than 40 percent of cloud deployments will leverage edge computing. Analysts say AI services will be among the earliest functions distributed across cloud and edge platforms, with 25 percent of endpoints that go live by the end of 2022 running AI algorithms.