Dominic Jainy is a seasoned IT professional with a deep specialization in high-performance computing, artificial intelligence, and the evolving infrastructure that supports them. As the digital landscape shifts toward decentralized, high-density power requirements, his expertise in integrating complex technologies like blockchain and machine learning into physical environments has made him a leading voice in data center strategy. Today, he shares his perspective on the ambitious transformation of the former Dumarey Powerglide factory in Strasbourg into a cutting-edge AI hub, a move that signals a new era for European digital sovereignty.
In this discussion, we explore the logistical complexities of repurposing massive industrial sites and the strategic importance of the FR-SXB1 project within the European Union’s broader supercomputing goals. Our conversation covers the innovative use of modular construction, the integration of biomass energy systems to meet staggering power demands, and the collaborative efforts between hardware and real estate giants to build a self-sustaining technological ecosystem.
The shift from an automotive gearbox factory to a high-tech AI data center is a massive leap in terms of utility. How does repurposing a 90-hectare industrial site like the Dumarey Powerglide facility change the traditional approach to building data center infrastructure?
Repurposing a site of this scale is as much about honoring the industrial past as it is about engineering the future. The sheer vastness of the 90-hectare footprint in Strasbourg’s Port du Rhin district provides a level of space that is rarely available for urban data centers, but it requires a complete rethink of internal logistics. Instead of the heavy, rhythmic clang of vehicle gearbox assembly, we are introducing the high-frequency hum of AI server racks and sophisticated cooling systems. Aether is utilizing the expertise of 2CRSi to create an on-site industrial facility specifically for designing and assembling prefabricated modules. This allows the team to build, test, and qualify data center units right on the factory floor before deploying them across the campus, which drastically cuts down on the typical construction timeline and logistical headaches of a 2027 opening.
With AI workloads requiring immense amounts of energy, how is the project balancing the immediate need for 25MW of power with the long-term sustainability goals of the European AI gigafactories program?
Securing 25MW of grid power for the FR-SXB1 site is a strong starting point, but the true innovation lies in how the consortium plans to supplement that energy. By integrating a biomass production plant directly onto the site, they aren’t just consuming energy; they are creating a circular system that supports the local environment. This plant will generate power and heat, but it also produces biochar, a carbon-capturing product that has a gritty, charcoal-like texture and serves as a vital soil enhancer for local farmers. When you consider that their total goal across two sites is 42MW, this self-sustaining energy model becomes a necessity rather than a luxury. It transforms a high-consumption AI cluster into a facility that actually gives back to the Strasbourg region through both carbon sequestration and agricultural support.
The project is part of a larger €20 billion EU initiative to build supercomputing clusters. How does the collaboration between firms like SiPearl, 2CRSi, and Dassault Systèmes ensure that this facility meets the high standards of European digital sovereignty?
This is a “dream team” of European technology, where each partner brings a specific, localized strength to the table to ensure we aren’t relying on external dependencies. You have SiPearl focusing on the high-performance chips that will fill these racks, and 2CRSi handling the intricate server assembly and module integration. By aiming to house 100,000 AI chips per cluster, they are building the raw muscle needed to train the most complex models known today. This isn’t just about competing with other consortiums like those led by Iliad or Volt; it’s about establishing a secure, indigenous supply chain for data. When the site eventually becomes Europe’s largest industrial hub for AI infrastructure, it will prove that we can build world-class compute power while maintaining complete control over the hardware and software stack.
What is your forecast for the development of AI gigafactories in Europe?
I expect that by the time FR-SXB1 and FR-SXB2 are fully operational with their combined 42MW capacity, we will see a rapid shift toward these modular, energy-integrated “gigafactories” as the standard. The success of the Strasbourg project will likely trigger a wave of similar redevelopments across the continent, where old heavy-industrial zones are revitalized into digital powerhouses that provide both high-tech jobs and sustainable environmental outputs. We are moving away from isolated “data boxes” and toward massive, integrated industrial ecosystems that can support the next generation of 100,000-chip supercomputing clusters. This movement will be the foundation that allows Europe to finally achieve true digital sovereignty, matching the compute scale of any other global region.