Engineering Excellence Since 1968!
20 May 2026
Advanced Technology and Artificial Intellience
Designing Data Centers for Speed, Scale, and Energy Reality
By John Hauge
The rapid growth of AI and advanced computing is reshaping how data centers are designed and delivered. For developers and operators, the challenge is no longer just building data centers. It’ s building them in a way that aligns with power availability, evolving technology, and increasingly compressed schedules.
Speed To Market Starts in Early Design
The challenges to rapid data center execution in the current environment are well documented: power availability, labor shortages, and equipment lead times, to name a few. Many of the risks that impact schedule and cost don’ t show up during construction— they originate in early planning.
Decisions made before procurement begins can lock in constraints around equipment, infrastructure, and delivery timelines. As a result, project teams are placing greater emphasis on early coordination. How can power be procured earlier? How can a design allow more labor to move from the field to a shop? What components can be used that can keep equipment procurement off of the critical path? These questions are being asked earlier and earlier in design processes as the industry drives toward faster execution.
This shift is helping teams better align construction timelines with market targets by making design decisions based on a construction process and not just a collection of lowest cost equipment pieced together by a collection of trades. In a market where delays can carry significant financial impact, that early alignment is becoming a competitive advantage.
Power Availability is Driving Project Strategy
In many regions, power availability is driving data center development. The issue is simple: Can the site get enough power, and can it get it in time? That constraint is reshaping the regions where these data centers can be built and how
RETAIL | COMMERCIAL | INDUSTRIAL | INSTITUTIONAL | MULTI-RESIDENTIAL
Engineering Excellence Since 1968!
THE VILLAGE ON TECHNOLOGY HILL
Rendering Courtesy of Built Form, LLC
DERRY, NH
Voted BEST NH Engineering Firm 14 Years Running!
LONDONDERRY, NH
ABBOTT COURT
Contact: Robert Duval, PE, LEED AP- President or Paul Sbacchi, PE- Chief Structural Engineer TFMoran, Inc. Bedford, NH and Portsmouth, NH |( 603) 472-4488 | www. tfmoran. com teams plan grid connections, manage loads, and build in capacity for future growth. In some cases, this leads to a more integrated approach that considers on-site generation, energy storage, or phased infrastructure deployment as part of the initial design.
Designing for Technology Change
One of the defining characteristics of today’ s data center market is how quickly the underlying technology is evolving. Chip density continues to increase, and new hardware generations are emerging at a pace that can outdate design assumptions before a facility is operational. Rather than attempting to predict exactly what future requirements will be, many teams are focusing on creating adaptable infrastructure.
Current predictions indicate that power density will continue to increase, suggesting that future challenges will no longer revolve around space within a data hall but instead around the ability to densify heating and cooling within the space. Where allowances used to be made in space programming for rack growth, those flexibility provisions are being shifted to electrical rooms and pipe galleries. That flexibility allows facilities to evolve alongside the technology they support.
Cooling Strategies are Evolving with Technology
Advances in chip design are also changing how data centers approach cooling. Newer technologies can operate effectively at higher temperatures, which opens the door to alternative cooling strategies, including reduced reliance on traditional compressor-based systems, allowed through the increased use of liquid cooling.
As the industry grapples with power limitations, maximizing IT computing capacity has become critical. Owners and designers are pushing for opportunities to leverage elevated water temperatures afforded by the latest AI technology to reduce the system peak PUE, which in turn allows for more of the site power to be used for IT load rather than supporting infrastructure.
These approaches can improve energy performance, but they also introduce new considerations. Highdensity environments require careful planning to maintain thermal stability, particularly during short-duration power events. To address this, design teams are leveraging increasingly advanced modeling techniques to better quantify and optimize loop temperatures during short breaks to avoid and minimize expensive supplemental systems like thermal storage.
Aligning Design with Owner Priorities
Data center projects vary widely depending on the owner’ s business model— whether hyperscale, colocation, or enterprise. Each comes with different priorities, from maximizing compute density to accelerating deployment or optimizing long-term operating costs.
Engineering teams are working more closely with owners to align design decisions with business outcomes, translating goals such as increased capacity or improved efficiency into practical infrastructure solutions. In a fast-moving market, that alignment is key to delivering facilities that not only come online quickly but continue to perform as demands evolve.
John Hauge, PE is a principal at Salas O’ Brien. He can be reached at john. hauge @ salasobrien. com. www. high-profile. com