The Data Center Frontier Executive Roundtable features insights from industry executives with lengthy experience in the data center industry. Here’s a look at the insights from Peter Panfil of Vertiv.
Peter Panfil is the Vice President of Global Power at Vertiv. He leads strategic customer development for the Power business and works to apply the latest power and control technology to proven and emerging topologies to provide the availability, scalability and efficiency levels customers demand. Approaching 30 years in critical space he has held executive positions including VP Engineering and VP/GM AC Power prior to his current responsibilities. He is a frequent presenter and spokesperson for industry trade shows, conferences and media outlets serving the IT, facilities and engineering industry.
Here’s the full text of Peter Panfil’s insights from our Executive Roundtable:
Data Center Frontier: The semiconductor shortage is among the examples of how the supply chain can impact business timetables. How would you assess the state of the data center supply chain, and how it has impacted the industry?
Peter Panfil: The supply chain issues that resulted from the pandemic are being addressed, but it will take time for the necessary capacity to come online. In the interim, we need to work together to buy ourselves that time. From a manufacturer’s perspective, we integrate a lot of chip-based intelligence into our systems, so we’ve had to work closely with our suppliers to understand how component and material availability would affect production schedules and new product introductions and then proactively communicate with our customers so there are no surprises.
Everyone is aware of the issues, which are global and not limited to our industry, so planning, communication and collaboration become the keys to making the best of a difficult situation. From an operator’s perspective, it’s more important than ever to make investments that can extend equipment life, such as preventive maintenance and remote monitoring.
Data Center Frontier: Recent weather-related disasters – including wildfires, winter storms and hurricanes – have tested long-held assumptions about power reliability and air quality in major data center markets. Will this “new normal” require changes in how the data center industry approaches site selection or facility operations?
Peter Panfil: Climate change has increased the severity and frequency of the disasters data center operators traditionally planned for, and that is already driving changes in design and operation. I’ll give you two examples.
Direct and indirect evaporative cooling systems became popular because they offered good energy efficiency. But water availability and air quality issues have reduced their appeal. The trend now is toward pumped refrigerant systems that offer energy efficiency similar to indirect evaporative systems while protecting data center systems from outside contaminants and supporting water-neutral sustainability goals.
For data center backup power, operators are re-evaluating their systems to ensure they can operate in normal, short-duration utility outages plus longer duration outages in which IT loads need to be shifted to other sites. Some are even exploring the feasibility of continuous operation without utility power. Moving to a continuous backup power source affects not only system component design, but also fuel considerations and concurrent maintenance. This has highlighted the need for locally generated power to be included in sustainability and business continuity plans.
Data Center Frontier: If “software is eating the world,” what does that look like for the data center industry? What are the opportunities and challenges presented by a world of “software-defined everything”?
Peter Panfil: The opportunity for the industry is continued growth by delivering the reliability and agility software-defined applications require. The challenges are in being able to dynamically manage changing loads within a facility and across facilities and building out the network edge, so software defined applications have the compute and connectivity they require.
We need to continue to compress deployment times for regional and local compute and storage through increased standardization and integration designed to the specific requirements of various edge use cases. We are also being asked to help increase the utilization rate of deployed assets. That has resulted in our product teams rerating our existing UPSes to deliver more kVA from the same platform and develop features that help customers overcome limitations of other parts of the critical infrastructure.
Data Center Frontier: As sustainability takes on greater urgency, hyperscale operators are testing new approaches to power infrastructure, including integrating more renewables and innovations in backup power and fuel. What are the most likely advances ahead in data center power?
Peter Panfil: Hyperscale operators have set ambitious sustainability goals. To realize those goals they are driving innovation in how data centers are powered, ultimately moving to locally generated renewable power supporting clean hydrogen production. The key barriers today are issues related to hydrogen costs, distribution and production, but progress is being made on all fronts. Advances in these areas, combined with the continued maturity of fuel cell and renewable energy technologies, will enable a solution in which renewably powered on-site hydrogen production supports fuel cells that serve as the primary source of data center power, with the utility, UPS and lithium-ion batteries working together to provide backup power as needed.
The traditional role of the UPS is expanding from power protection and conditioning to include demand management, fast frequency response and grid services.