POWER


is entering an era where data centres are not built but assembled. Modular prefabricated data centre manufacturing is becoming increasingly common in an environment where, only a few years ago, such approaches were largely confined to niche edge deployments. Modular electrical rooms - integrating UPS, switchgear, batteries, control systems and monitoring into unified, factory-assembled units - are emerging as a preferred solution for AI- ready infrastructure.


By consolidating critical components into a single engineered environment, modular power architectures can provide a more stable foundation for handling the volatile demands of AI workloads while also enabling faster, more predictable deployment. Modularity also offers a practical response to the growing compression between infrastructure lifecycles and the pace of AI hardware evolution. Power systems have traditionally been designed around multiyear planning cycles, while AI compute platforms may be refreshed annually or faster, increasing the risk of misalignment at commissioning. Modular architectures perserve flexibility, allowing power blocks to be replaced, reconfigured, or scaled as requirements evolve. Rather than locking operators into monolithic designs based on shortlived assumptions, modular systems support adaptation under uncertainty, redefining resilience as the ability to evolve without compromising reliability or asset life.


New responsibilities upstream for the supply chain


The rise of prefabrication and modular deployment is shifting greater system- level responsibility upstream within the supply chain. Increasingly integration and testing take place in factory environments, requiring suppliers to provide pre-validated subsystems rather than standalone components.


This shift has several implications: • • •


Suppliers must design for integration, ensuring compatibility across electrical, thermal and control interfaces


Factory-level testing becomes critical, reducing the need for on-site commissioning and rework


Earlier and deeper collaboration is required, including closer engagement with chipmakers and system integrators


At the same time, traditional industrial equipment supply chains are under significant strain. Demand for critical components such as transformers and medium-voltage switchgear has surged, leading to extended lead times and capacity constraints. These challenges expose the limitations of legacy manufacturing and procurement models, which were designed for more stable and predictable markets. While this shift enables faster and more repeatable deployment, it also places greater responsibility on equipment providers and ecosystem engineering partners. In a modular world, expectations


increasingly centre on delivering systems that integrate cleanly with equipment from other suppliers, function reliably as part of a wider architecture, and require minimal modification during installation.


Conclusion


AI has already changed the face of the internet. It has changed the way millions of people work. And it is now driving equally profound change in how data centres are designed, built and powered.


AI workloads are more powerhungry, less predictable and are evolving faster than traditional computing, placing sustained strain on power infrastructure and accelerating wear on equipment such as switchgear and UPS systems.


Managing these challenges can no longer be done solely within the data centre itself. It increasingly requires closer coordination upstream, including how equipment is designed, manufactured and integrated. As modular construction and offsite prefabrication become more common, greater responsibility for system reliability, compatibility and longterm performance shifts toward suppliers and engineering partners.


Ultimately, the industry’s ability to support the AI boom depends heavily on power. Meeting this challenge - across both demanding AI workloads and modular build models - will require power systems that are not just strong, but able to adapt as conditions and requirements change.


JUNE 2026 | ELECTRONICS FOR ENGINEERS 19


Page 1  |  Page 2  |  Page 3  |  Page 4  |  Page 5  |  Page 6  |  Page 7  |  Page 8  |  Page 9  |  Page 10  |  Page 11  |  Page 12  |  Page 13  |  Page 14  |  Page 15  |  Page 16  |  Page 17  |  Page 18  |  Page 19  |  Page 20  |  Page 21  |  Page 22  |  Page 23  |  Page 24  |  Page 25  |  Page 26  |  Page 27  |  Page 28  |  Page 29  |  Page 30  |  Page 31  |  Page 32  |  Page 33  |  Page 34  |  Page 35  |  Page 36  |  Page 37  |  Page 38  |  Page 39  |  Page 40  |  Page 41  |  Page 42  |  Page 43  |  Page 44  |  Page 45  |  Page 46  |  Page 47  |  Page 48  |  Page 49  |  Page 50