Powering the AI revolution: How data center ATS requirements are reshaping OEM designs
Cooling systems have become mission critical as the need for data center capacity surges. Incorporating modern ATS technology into data center designs can help speed deployment and cooling system reliability, giving OEMs a clear, defensible competitive advantage.
Relentless demand for digital services, driven by the AI revolution, is forcing data centers to operate at the edge of what their electrical and mechanical infrastructure can support. Instead of single, isolated constraints, operators now contend with intertwined technical, commercial, and operational forces that amplify one another. The result is an environment where design decisions, technology choices, and deployment strategies must be reconsidered from the ground up to sustain performance, uptime, and profitability.
The perfect storm: Three converging challenges
Explosive AI adoption is transforming data centers from a steady-growth utility into one of the fastest-expanding infrastructure asset classes. Recent analyses estimate that demand for AI-ready data center capacity will rise at about 33% per year from 2023 to 2030. By mid-decade, AI is expected to account for roughly 30% of total data center power demand, up from about 14% in 2023. By 2030, this growth will push global capacity toward 170–220 GW, nearly doubling total data center electricity use.
As the AI‑driven buildout accelerates, the limiting factor is not just how much capacity operators can add, but also whether their power, switching, and cooling infrastructure can keep pace. Three challenges are converging to create a perfect storm:
- Speed-to-market pressure is intensifying – Data center operators are racing to deploy capacity ahead of competitors. Construction timelines that once spanned 24-36 months are now compressed to 12-18 months.
- Density demands are skyrocketing – AI and high-performance computing applications are pushing power density to levels previously unimaginable. Traditional data centers operated at 5-10 kW per rack, but AI-optimized facilities now require 30-100 kW per rack, with some hyperscale deployments exceeding 200 kW.
- Cooling infrastructure has become mission critical – If a Computer Room Air Handler (CRAH) and Computer Room Air Conditioning (CRAC) unit drops offline, temperatures can spike in minutes, so maintaining power to cooling systems is now just as vital as feeding the servers themselves.
To address the increased mission-critical nature of cooling systems, many modern data center designs call for dedicated ATS protection for each cooling bank. In practice, this means a single 3 MW facility may require 15–20 ATS units just for cooling infrastructure, turning ATS installation into a critical constraint for energizing new capacity. When every additional megawatt must be brought online faster and packed into denser footprints, the way ATS devices are designed, installed, and& maintained becomes a key lever for both faster deployment and operational resilience.
How intelligent ATS designs accelerate data center deployment
As data centers scale from a few switches to dozens or hundreds per site, modern ATS platforms make high‑density growth manageable by dramatically simplifying installation, commissioning, and service. Integrated designs consolidate separate controllers and dense wiring into compact, all‑in‑one assemblies, which can cut installation time from roughly 8 hours to about 90 minutes per unit and free up more than 150 labor hours on a 20‑switch project. That time savings helps projects energize weeks sooner, supporting earlier revenue recognition and giving OEMs a more compelling value story without adding complexity for their build teams.
Scalability and standardization
Contemporary ATS architectures are also engineered for repeatability and scalability, which is critical when cooling, IT, and power systems must expand quickly without sacrificing reliability. Modular, serviceable components enable fast replacements in minutes instead of hours, helping operators maintain stringent uptime SLAs even as the number of ATS units on site grows. For OEMs, this means fewer custom one‑offs and more standardized, application‑specific designs that are easier to reproduce, stock, and support over time.
Intelligence and remote monitoring
Intelligence built into modern ATS solutions further elevates their role from simple transfer devices to active reliability assets for the facility. Features such as predictive maintenance, remote diagnostics, and centralized monitoring allow operators managing fleets of 50 or more ATS units to reduce site visits, schedule maintenance proactively, and prevent issues before they disrupt cooling or IT loads. This connected visibility is especially valuable in distributed or rapidly expanding data center portfolios, where traditional, manually managed systems can no longer keep pace.
ATS solutions
ATS offerings from vendors like ABB illustrate how modern ATS devices deliver immediate, tangible returns. For example, ABB’s TruONETM ATS integrates controller and switch functions to reduce wiring and installation time by up to 80%. It also supports application-specific CRAH/CRAC configurations across common data center voltage and current ranges.
In addition, cloud-connected diagnostic tools provide centralized monitoring, predictive maintenance, and remote troubleshooting for fleets of ATS devices. ABB’s Lite Panel Pro can remotely monitor up to 90 devices, including ATS, from a single unit. For a facility managing multiple ATS units, this level of visibility can cut site visits by up to 40% while helping operators detect issues sooner and resolve them faster.
The path forward
The data center boom fueled by the AI revolution isn’t slowing down, so neither is the demand for capacity. With power distribution failures responsible for 54% of all impactful data center outages and downtime costs averaging $9,000 per minute, the ATS is no longer a commodity component — it is a strategic asset. By incorporating components that are faster to install, easier to scale, and smarter to operate into their data center solutions, electrical OEMs can turn speed-to-market pressures and extreme density requirements into a competitive advantage. Modern ATS devices provide a clear path to delivering the resilient, intelligent switching backbone needed to bring new AI capacity online quickly and keep it running reliably.
—
Babu Chinnasamy, Product Marketing Manager
