Orbis Electric’s Deep Automotive Experience Aids its Data Center Cooling Future

How Orbis Electric’s Deep Automotive Roots are Powering its Data Center Cooling Future

By Marcus Hays, founder & CEO of Orbis Electric

When Orbis began exploring data center cooling, it wasn’t chasing a trend – it was recognizing a pattern.

In electric vehicles, electric trucks, and electric refrigeration units (eTRUs), Orbis has spent more than a decade solving the same core problem data centers now face: how to move, convert, and manage energy with extreme efficiency under real-world constraints. Mobile electrification doesn’t allow for waste. There’s no luxury of excess power, no tolerance for inefficiency, no oversized footprint, and no patience for unnecessary heat.

Until recently, data centers operated under very different assumptions. But those assumptions are disappearing — fast.

As AI workloads push GPUs to unprecedented power densities, cooling has become one of the largest – and least optimized – energy loads in modern data centers, consuming as much as 30–50% of total facility power. Yet much of today’s cooling infrastructure still relies on legacy architectures: large three-phase AC induction motors, multiple conversion steps, bulky variable frequency drives, and supply chains stretched thin by hyperscale demand and geopolitical constraints.

To Orbis’ engineers, this landscape felt strikingly familiar, which ultimately led to the development of the Cooling Engine, powered by Orbis’ HaloDrive™ axial flux motor platform.

Liquid cooling, whether glycol-based or dielectric, is how electric vehicles have always protected batteries and power electronics. Managing DC power efficiently, minimizing conversion losses, and designing compact, high-performance systems is core to Orbis’ DNA. In mobile applications, there’s no room for band-aids. Everything must be purpose-built for efficiency from the start.

Just as importantly, automotive electrification forced Orbis to perfect propulsion-grade motors capable of delivering exceptionally high torque at low rotational speeds. That capability – essential for moving vehicles and refrigerated trailers reliably – turns out to be equally critical for pumping large volumes of coolant through dense, high-power chip stacks. In data centers, torque translates directly into stronger, more responsive pumping performance, resulting in fewer energy losses.

Rather than layering new components onto old architectures, Orbis approached data center cooling holistically – designing a DC-native system that strips out unnecessary conversion steps, eliminates traditional standalone variable frequency drives, and dramatically reduces energy loss. The result is a 2MW, OCP-compatible coolant distribution unit with up to a 50% smaller footprint than comparable systems, and efficiency gains that compound at scale.

Equally important is how Orbis builds. 

Automotive electrification forced Orbis to develop technology that could be manufactured at scale for the global automotive supply chain. That same infrastructure is now uniquely positioned to serve a data center market facing acute shortages and long lead times. While others are constrained by limited industrial capacity, Orbis can move faster because it already knows how to scale hardware under pressure.

So, why does this matter now? Because AI has changed the equation for data centers. Cooling is no longer a background system – it is one of the primary constraints on how fast compute can scale.

In many ways, the Cooling Engine represents the same transition electric vehicles once made: from brute-force solutions to systems engineered for efficiency, integration, and sustainability. At this inflection point, efficiency isn’t just about saving energy – it’s about unlocking headroom for more compute power.

Orbis Electric’s years in automotive electrification have served as an invaluable proving ground for what comes next – the great AI cooling challenge.