There is good reason that Google opted to apply artificial intelligence in its bid to reach the lowest PUE in the industry. The mechanical system of any data center represents the low-hanging fruit on the tree of cost savings through energy efficiency. Their oft-cited project produced a 40% reduction in operating costs associated with cooling. What may not be as apparent is the role that data center PDUs play in this story.
But first, let’s talk about cooling.
Cooling systems, as pointed out in this article by Alan Seal on vXchnge.com, are incredibly complex. In order for them to operate optimally, there must exist a direct line of communication between the innumerable servers and computing devices that are generating the load. Humans can help with manual load balancing but completing that task in real time to scale is simply too cumbersome and often complex for even the savviest team of engineers. In short, there is simply too much data to take in and process, and too many decisions that need to be made.
With that said, a cautionary note before we move headlong into artificial intelligence. As mentioned in this article in The ACHR News: “AI will likely impact all aspects of HVACR equipment and controls, but people should not expect magical results overnight…as end users will see the true value of AI over time, with benefits that include improved comfort, energy savings, asset life, and predictive maintenance.” It is important to focus on what you can and cannot improve with AI, and it is likewise important to set expectations before tackling these kinds of projects. Artificial intelligence is not the be all, end all.
So why all of the talk about AI and data center cooling? Undoubtedly, it yields operating efficiencies. Primarily, there’s a reduction in power consumed by the mechanical system, but also less wear and tear on those systems. The secondary benefits are less obvious.
Optimizing the operation of the mechanical infrastructure has several hidden cost benefits. First, it could lead to a reduction in the amount and regularity of required maintenance on systems. Second, equipment that is operated efficiently and within optimal conditions tends to last longer, which means that the useful service life of key components such as pumps and chillers can be extended. Pushing replacement timeframes out into the future extends the parameters of the financial model. The net effect is a reduction in the total cost of ownership.
What is the role of the intelligent PDU in the data center environmental monitoring struggle? First, rack PDU(s) are uniquely positioned to provide cost-effective data collection points at the rack. Along with utility-grade power monitoring, they can also be leveraged to collect temperature and humidity points back to the system in real time. Valuable insight can be delivered by smart sensors that provide real-time environmental alerts, notifying you immediately when something is astray, decreasing the response time to remediate any critical risks. This combined power distribution, management, and monitoring approach leverages the investment of rack PDUs. It makes them the eyes and ears, so to speak, within the rack. Having an intelligent PDU is like having thousands of thermostats in your facility, right where you need them so you can easily recognize hot spots, optimally cool equipment, extend equipment life and prevent costly downtime.
To manage cooling costs and leverage the benefits of AI in your data center, you can start with intelligent PDUs and turn data about your data center into actionable information, and from there you can realize cost savings and increased efficiency for your bottom line.
About the Author
Rebecca Gilstrap is Director of Strategy for Legrand’s Data, Power and Control Division. In this role she leads cross functional teams to set, prioritize and achieve strategic growth initiatives. Rebecca has spent the last decade optimizing data centers and mitigating IT infrastructure outages. With a career rooted in security compliance and business continuity, she has worked across organizational boundaries to highlight critical IT dependencies. She has managed international technology divestiture projects, built colocation businesses from the ground up and created multi-tenant cloud service platforms.