6 Data Center Trends You Should Know
It is an AI and cloud-based world. That’s another way of saying that these two transformative technologies will overhaul data center architecture.
Few technologies have shaped business in the past 30 years as much as the cloud and its enabling technology, mobile connectivity. Business data and IT infrastructure have moved off-premise and into data centers, whether public or private, mirroring the same trend among consumers.
This has shaken up the data center landscape, and also made the data center more important. Semiconductor major Nvidia that created an addressable market for GPUs and gaming is shaping up its data center strategy with Mellanox acquisition.
The data center is not standing still, though. Roughly 80% of enterprises will shut down their traditional data centers by 2025 in favor of hybrid or public cloud setups, according to Gartner research, and those that remain will not look the same.
Here are six data center trends you should know, from broad trends to specific changes going on within data center design right now.
Learn More: How to Successfully Mobilize Data Center AI Applications while Avoiding Business Disruption
1. Hybrid Architectures on the Rise
“While we still run into customers that want to move it all to public cloud, we are starting to see that customers that moved aggressively to public cloud are starting to pull some workloads back into their own on-premises or colocation data centers,” says Tom Johnston, principal architect for consulting services at cloud and data center transformation advisory, Insight.
Some of this is from businesses not understanding their workloads fully, according to Johnston, and part is from unrealized cost savings expectations.
Solutions that extend the consumption of services from the public cloud providers into the on-premises environment are becoming a trend as a result, he notes.
“For example, with Microsoft’s Azure Stack family of solutions, there are many ways to provide Azure cloud-based services to end users while also retaining data and services on-premises,” he says. “Amazon AWS recently released a similar offering known as AWS Outposts.”
2. Less Specialization
Data Centers are transitioning away from specialized workloads and partitioned environments. Instead, modern data centers are embracing a virtualized, single software fabric that runs all workloads. This includes SQL databases, virtual desktops, big data and analytics services, containers with Kubernetes, and others.
Data center use is getting more sophisticated as the center of gravity shifts from individual data centers to applications that stretch across multiple clouds. A web tier might reside in the cloud, for instance, with databases housed on-premise.
3. Immersion Cooling for Better Heat Management
Sustainability considerations such as cooling and energy efficiency continue to play an outsized role in the data center, both for economic and environmental reasons.
“Operational efficiency is central to the environmental sustainability of the data center,” notes Zack Zilakakis, a former Gartner research director now with data center automation provider, Apstra.
One move in that direction is toward immersion cooling.
“This method of cooling is slowly becoming popular as hardware manufactures and infrastructure manufactures are starting to develop this leading-edge cooling topology,” says Phil Rafferty, president of data center design and construction firm, Data Specialties Inc.
“With Immersion cooling, the computer components are submerged into a dielectric coolant eliminating the need for fans in the IT hardware and large cooling units to remove the heat from the room.”
4. Added Security Everywhere
To protect against a wider array of threats, both internal and external, data centers are taking extraordinary steps to encrypt all data, firewall individual applications and services with micro-segmentation, and apply fine-grained access controls for every part of data center operations.
“With more cloud infrastructures emerging, security is a major design trend for all data centers today,” notes Kamran Amini, vice president and general manager of server, storage and software-defined infrastructure at Lenovo.
Learn More: New Tech Product Launches in AI, Data, Cloud and more: Week of June 8, 2020
5. Automating Operations
As with much of business, machine learning and automation also are picking up within the data center.
Data center architects are starting to use machine learning techniques that apply precise algorithms to analyze historical data and make detailed recommendations about data center adjustments. By combining this with automation, data centers are moving toward self-optimization where capacity additions, application provisioning, troubleshooting and other parts of data center operations are configured and reconfigured automatically.
Having a single virtual fabric used for allocating workloads in the data center helps accelerate this trend, obviously.
“Automating the implementation and operations of the data center makes managing the infrastructure more repeatable and less error-prone,” says Zilakakis.
6. Colocating for Other Companies
Finally, an interesting data center trend quietly emerging is companies renting out part of their data center space to other firms, effectively becoming a colocator.
“While it’s strange for enterprises to rent out part of their data center to other companies, it’s really a real-estate transaction, not unlike subletting office space that is no longer needed,” says David Linthicum, managing director and chief cloud strategy officer at Deloitte Consulting. “There are cost advantages.”
These are interesting times for the data center, and its evolution looks set to continue for the foreseeable future.
Is the data center market headed for a slowdown? Comment below or let us know on LinkedIn, Twitter, or Facebook. We’d love to hear from you!