Submer Group, the full-stack AI infrastructure company, has announced the launch of Corenix, a new modular data center infrastructure company focused on accelerated computing. Headquartered in Houston, Texas, Corenix will serve neoclouds, hyperscalers, and AI operators with custom-engineered modular data center platforms based on NVIDIA reference designs. The company has appointed former Johnson Controls executive Martin Renkis as CEO to lead the new venture.
Corenix designs, builds, integrates, and factory tests complete functional data center modules covering IT, power, cooling, and networking at production facilities before delivering them to project locations. Installation, connection to site infrastructure, site acceptance testing, commissioning, and operational readiness then take place at the client site.
Corenix manages the agreed project scope across both stages, bringing engineering, manufacturing, supply chain coordination, and on-site deployment together into one delivery model. Each Corenix platform is based on an NVIDIA reference design and custom-engineered to the client's own requirements.
Traditional data center builds cannot keep pace with the power, networking, cooling, and serviceability demands of accelerated computing, or with the need to support successive generations of high-density compute. Corenix was built to close that gap with a repeatable path from an approved technical architecture to operational capacity at the required location.
Corenix draws on Submer Group's experience in high-density thermal engineering, global infrastructure deployment, and operations, proven at scale by a campus with hundreds of megawatts of capacity delivered as a fully modular compute buildout in nine months from order to delivery, and over a decade of liquid cooling engineering.
"AI factories are driving unprecedented demand for new data center capacity, and the market needs a faster, more predictable way to build it," said Patrick Smets, CEO of Submer Group. "Corenix builds on Submer's history in modular data center delivery and the engineering, thermal, and infrastructure expertise we have developed solving high-density compute challenges for years. We are bringing that experience into a dedicated business built to deliver AI infrastructure at scale."
"Customers need clarity before they commit: what needs to be built, how it will be tested, who is accountable, and when it will become operational," said Renkis. "That's the foundation I'm building Corenix on from day one. Corenix puts a dated program and a go-live date on the table before a contract exists. One team stays accountable from factory integration and testing through on-site commissioning and operational readiness, with the same team offering the flexibility to support current and future accelerated computing systems."
Renkis joins Corenix from Johnson Controls, where he served as executive director of Data Center Infrastructure Services and developed AI infrastructure offerings covering power, cooling, heat recovery, operations, and financing.
Corenix works alongside two other Submer Group companies where their capabilities support a customer's full stack AI infrastructure deployment requirements: Rubix Data Centers, which develops and operates AI data center campuses, and Radian Arc, which provides GPU-as-a-Service and AI-as-a-Service.
About Corenix
Corenix, a Submer Group company, designs and delivers industrialized, customizable modular data center platforms based on NVIDIA reference designs. Corenix is focused on the deployment layer of AI infrastructure, delivering rack power, cooling, networking, and IT integrated and factory tested as one system, to bring AI capacity online against a fixed date. Corenix is headquartered in Houston, Texas, U.S.A.
About Submer Group
Submer Group brings together specialist operating companies across the AI infrastructure stack to close the gap between AI ambition and AI reality. Built on a decade of infrastructure engineering, the group now spans from land and power to cloud and edge. The group is built to address the structural constraints of AI infrastructure, namely power availability, thermal density, deployment speed and data sovereignty, enabling AI to be deployed efficiently and at scale. The group has deployed 500 MW+ of liquid-cooled infrastructure and holds 8 GW+ of powered land across the Americas, EMEA and APAC. Founded in Barcelona in 2015, the group operates globally, with hubs in Houston and Taipei.
