From Optional to Essential: DLB Uses Digital Twins to Engineer AI Data Centers

AI Is Changing the Rules of Data Center Design

Artificial intelligence is reshaping the data center into one of the most complex engineering environments in the built world. As rack densities rise, liquid cooling adoption accelerates, and power and thermal constraints become increasingly interdependent, traditional design methods alone are no longer sufficient. For the next generation of AI infrastructure, digital twins are moving from advanced engineering tools to mandatory essential platforms for design confidence, risk reduction, and operational readiness.

Decades of Experience in Mission-Critical Infrastructure

That shift is central to the work of DLB, a long-standing engineering and consulting leader in mission-critical data center design. Founded in 1980, DLB has built a strong reputation across the industry for data center site selection, due diligence, design engineering, commissioning, construction quality management, and energy optimization. The company has supported leading hyperscalers, emerging hyperscalers, neo-clouds, enterprises, and colocation providers, helping customers accelerate speed to market while navigating the complexity of mission-critical infrastructure.

In January 2026, Accenture acquired a majority stake (65%) in DLB, and the DLB team’s joined Accenture’s Industry X practice, expanding end-to-end data center capabilities from early-stage site development and conceptual design through engineering, deployment, and operational performance.

Designing AI Infrastructure Before It Is Built

To help customers address the challenges of next-generation AI infrastructure, DLB is leveraging the Cadence Reality Digital Twin Platform on high-density AI projects Cadence Reality enables teams to create engineering-grade digital representations of data centers and campuses that connect compute, cooling, power, airflow, facility systems, and environmental conditions into a single simulation environment. This chip-to-campus approach, which extends simulation from the chip and rack level through the data hall to the surrounding campus, provides visibility into how systems interact under real-world operating conditions, enabling predictive analysis, capacity planning, energy optimization, and infrastructure planning before physical deployment.

For DLB, this means customers can evaluate performance, efficiency, and scalability earlier in the project lifecycle, when changes are easier and less costly to make.

“The industry is moving from low-density data centers to high-density AI infrastructure, and that changes the engineering process,” said Nick Gmitter, Design Director at DLB. “Digital twins and software platforms like Cadence Reality Digital Twin are becoming an essential part of the engineering workflow because customers need to understand how the full system will behave before they build. With Cadence, we can help customers virtually prototype high-density AI environments, evaluate power and cooling trade-offs, and make more confident infrastructure decisions before deployment.”

From AI Racks to Campus Systems

As compute densities increase, engineering decisions can no longer be made in isolation. The behavior of servers, airflow, liquid cooling, power delivery, cooling distribution networks, and site infrastructure must be evaluated as an interconnected system. By combining high-fidelity simulation with flow network modeling, DLB can help customers analyze trade-offs, validate cooling concepts, improve resource utilization, and understand system-wide impacts before construction or expansion.

The same perspective extends beyond the data hall. External weather conditions, cooling plants, power infrastructure, on-site generation and energy storage, and campus-level energy systems all influence operational performance. The platform’s chip-to-campus approach enables these dependencies to be evaluated within a single engineering framework.

Reducing Uncertainty Before Deployment

High-density AI facilities leave little room for trial and error. Virtual prototyping allows teams to compare alternatives, validate assumptions, assess operational readiness, and reduce uncertainty before major capital investments are made. Rather than discovering limitations after deployment, organizations can evaluate performance and resilience while the project remains flexible.

The New Foundation for AI Data Center Engineering

As AI infrastructure accelerates, the market is entering a new phase. Digital twins were once viewed primarily as tools for design validation and optimization. Today, they are becoming foundational to how modern data centers are designed, commissioned, and operated.

By combining DLB’s decades of mission-critical engineering expertise with the Cadence Reality Digital Twin Platform, customers gain a more predictive, simulation-driven approach to AI infrastructure design and lifecycle management. The result is a stronger foundation for designing high-density, efficient, resilient, and scalable data centers before design assumptions become physical constraints.

Learn more about Cadence Reality Digital Twin Platform and about DLB Associates.