Building AI-Ready Data Centers Means Designing For Certainty

Abhilash Kunnatoor Margabandu is vice president of infrastructure engineering at EdgeCore Digital Infrastructure.

getty

​For years, data center design has been grounded in predictability. Engineers could model around demand, define capacity and build facilities that would reliably serve their purpose for a decade or more. Today, with AI entering the scene and transforming the industry as we know it, data centers must be designed with flexibility while still providing customers with certainty around performance, timelines and security.​

Here are the aspects leaders must prioritize today to keep pace with industry shifts.

Starting At The Site Level

Constraints in power and land are happening at a pace the industry has never before seen. In this environment, certainty cannot come from assumptions. It must come from deliberate and foundational design choices rooted in flexibility. One of the biggest misconceptions about future-ready infrastructure is that adaptability begins only inside the facility. But in reality, it starts much earlier, with site selection.​

Land, power availability and network proximity are the variables that define what is ultimately possible on a campus. If those components are not properly understood and secured upfront, no amount of downstream engineering can compensate.​

Today, it’s all about selecting sites that can not only support current demand but also multiple generations of technological evolution. That includes planning for significantly higher power densities, evaluating long-term grid capacity and, in some cases, incorporating on-site or adjacent generation like natural gas to ensure reliability and speed to market. In a world where power is increasingly the limiting factor, control over energy strategy is one of the clearest paths to certainty.​

Engineering For The Known Unknowns

While the exact trajectory of AI hardware is uncertain, the direction is not. We know that densities are increasing and speed to compute-ready is intensifying. We also know that power requirements are concentrating at the rack level. Designing with certainty means building systems that can handle these unknowns.​

Take cooling as an example. Liquid cooling has emerged as the new standard for high-density facilities. However, the specific ratio and form it will take over the next decade is still unclear. Future-ready designs don’t stick to a single approach; instead, they enable multiple cooling options ranging from air, liquid and hybrid, all found within the same facility.​

The same applies to electrical systems. Delivering significantly more power through smaller footprints requires rethinking everything from distribution architecture and equipment sizing to redundancy models. Operators must design for both modular upgrades and future capacity expansion to meet today’s needs and prepare for tomorrow. In this circumstance, certainty isn’t about focusing on one single solution, but ensuring infrastructure can support a range of potential outcomes.​

Finding Precision In Power Delivery

If there is one area where certainty cannot be compromised, it’s power. As densities increase, the tolerance for inefficiency and variability decreases. Even minor improvements in power usage effectiveness (PUE) and electrical efficiencies have a significant impact, especially in gigascale developments.​

The digital infrastructure industry needs to critically address the challenge of power availability. Grid constraints are slowing development across major markets, so integrating alternative energy strategies, like on-site generation and energy storage, adds another layer of certainty by reducing reliance on external sources.​

To remain successful, organizations must have a deep focus on every component of the electrical system, from transformers and breakers to distribution pathways and monitoring technologies. They also should shift away from designing for theoretical limits and instead base plans on realistic, evolving types of demand.​

Planning And Building Smarter

Speed has undoubtedly become a defining factor in data center developments. We’re seeing AI infrastructure demand outpacing supply, while hyperscalers are prioritizing partners who can deliver capacity quickly and reliably. But how can we accelerate timelines without introducing more risk?​

The answer is standardization and process innovation. Modular construction and closer collaboration with supply chain partners can help speed up builds while maintaining engineering precision. While supply chain challenges remain an unfortunate reality—influencing everything from equipment selection to site strategy—designing with certainty means accounting for these constraints early on and building contingency into both design and execution.​

Mastering The Human Factor

Success in today’s digital infrastructure largely depends on the people who plan, design, build and operate it. But today, the industry is facing a significant workforce gap.​ Even as demand increases, there’s a shortage of specialty roles like electricians, operators and experienced engineers. Adding to the challenge is the fact that the complexity of modern data centers requires a broader, more integrated skill set. Engineers must understand how all systems—mechanical, electrical and automation—interact with one another.​

Without the right talent, even the most carefully designed systems cannot deliver on their promise. Addressing the challenge means investing in training, upskilling and workforce development.​

Redefining Certainty Over The Next Decade

The concept of future-ready data centers is not obsolete, but it does need to be redefined. In today’s fast-moving landscape, it’s not about predicting every outcome, but building a foundation that can support continuous evolution without requiring constant reinvention.​

AI has introduced a new level of complexity into data center design, but it has also clarified what matters most: marching toward certainty. The developers who “win” over the next decade will be those that design with conviction, grounded in fundamentals and built to perform under a wide range of conditions.​

Forbes Technology Council is an invitation-only community for world-class CIOs, CTOs and technology executives. Do I qualify?