A multidisciplinary consortium led by Swiss-US energy company Deep Atomic is seeking to develop next-generation, nuclear-powered AI data centre and energy infrastructure campuses, including a proposed facility at the Idaho National Laboratory (INL) pending approval from the US Department of Energy (DoE)
The group combines expertise in nuclear engineering, data centre design and high-density construction. Lead developer Deep Atomic said in October 2024 that it had designed the MK60, light water small modular reactor (SMR) to serve the requirements of modern and future data centres.
According to Deep Atomic, the MK60 SMR is designed using Generation III+ pressurised light water reactor technology. The reactor’s balance of plant is designed to produce 60MW of electricity and 60MW of cooling.
The 60MW of electricity is generated by a steam turbine generator set utilising high-temperature and pressure steam from the nuclear steam supply system. The 60MW cooling plant is driven by low-grade steam from the turbine. The target date for the project (series manufacturing) is the end of 2029. In March 2025, Deep Atomic expressed its intent to engage with the US Nuclear Regulatory Commission (NRC) to start a pre-application process for the design certification of the MK60.
Other consortium members include:
- Clayco: A major design-build company tasked with the master planning, engineering and construction of the data centre facilities
- Future-tech: A leading data centre design and advisory practice that handles site master modelling and architectural due diligence
- Paragon Energy Solutions: Provides nuclear-grade components and supply chain expertise
- Moonlite: An AI infrastructure company focused on the specialised hardware requirements of high-density compute workloads.
The consortium is supporting development of the MK60, which leverages waste heat to provide direct liquid cooling for AI servers, significantly increasing efficiency. The facilities are designed to operate independently of the main utility grid, reducing strain on local power networks. The current proposal for the INL site suggests starting data centre operations within 24–36 months using existing grid and renewable power (solar/geothermal) while the MK60 reactor undergoes final certification and commissioning.
As part of the consortium, Clayco is supporting the data centre scope by providing early-stage delivery planning, constructability analysis and integrated design-build expertise to inform the DoE submission process. In this role, Clayco provides expertise on how the campus could be designed, engineered and constructed, from early site development strategy through construction phasing and sequencing.
“Successful DoE submissions require more than innovative energy concepts. They require confidence that projects can be delivered safely, efficiently and at scale,” said Bob Clark, executive chairman and founder of Clayco. “Our team is proud to support Deep Atomic and the broader consortium by bringing real-world data centre delivery experience and execution discipline to help present a buildable, delivery-ready proposal for this groundbreaking, first-of-kind project.”
Clayco will also advise the consortium on infrastructure integration and aligning design and construction approaches with the operational requirements of high-density AI workloads.
“Our mission is to accelerate the deployment of advanced nuclear power solutions that meet the performance, reliability and sustainability needs of next-generation computing infrastructure,” said William GJ Theron, founder and CEO of Deep Atomic. “Working with Clayco and our consortium partners, we are proud to present a proposal to the DoE that pairs practical execution planning with innovative energy and data centre integration.”
Deep Atomic is advancing an integrated energy approach that pairs advanced nuclear energy with AI data centre infrastructure to support reliable, low-carbon power for next-generation computing. Clayco’s involvement reinforces the consortium’s ability to translate this vision into a structured, executable campus concept aligned with DoE evaluation criteria.
