A new tender for an AI-focused high-performance data centre highlights how public bodies across Europe are upgrading infrastructure to run intensive AI workloads.
Follow Tenderlake on LinkedIn for concise insights on public-sector tenders and emerging procurement signals.
On 26th December 2025, SURF B.V. quietly set out to buy a new backbone for its artificial intelligence ambitions: the delivery of a high performance computing datacentre for its AI facility. The short contract notice signals a substantial move from experimenting with AI tools to commissioning the dedicated infrastructure needed to run them at scale.
The procurement describes the delivery of a high performance computing datacenter for the AI facility of the contracting authority. While the notice offers few technical details, the decision to tender an entire datacentre facility rather than a handful of servers is telling.
High performance computing (HPC) infrastructure is the unseen layer that makes it possible to train complex machine learning models, process large data sets and support intensive analytics. For a public organisation, investing in a dedicated HPC datacentre suggests an intention to embed AI into core services and research, rather than treating it as a side project.
The summary does not specify capacity, location or the balance between compute, storage and networking. Yet similar projects indicate the direction of travel. In October 2025, Vytauto Didžiojo universitetas set out to acquire a high-performance computing platform for accelerated data processing, storage and archiving, explicitly bundling delivery, installation, configuration and staff training. Buyers are not just purchasing hardware; they are commissioning operational environments.
The SURF B.V. tender lands in a year marked by a wave of public-sector investment in AI-focused computing across Europe. In July 2025, UNIVERSITATEA TEHNICA DIN CLUJ NAPOCA launched a server procurement covering a high-performance computing system for artificial intelligence, servers to collect TV and satellite streams, and development tools for processors. On the same day, Universitatea de Vest din Timisoara published its own advanced AI computing infrastructure notice, seeking processing nodes, data storage servers, interconnection switches and management platforms to create a high-performance ecosystem for machine learning.
AI compute is also moving into specialist public functions. On 16th July 2025, Serviciul de Telecomunicatii Speciale went to market for technological equipment and services to set up a high performance computing data room as part of a project to enhance forest monitoring and evaluation through an integrated IT system. In November 2025, the Polish national fire service, KOMENDA GŁÓWNA PAŃSTWOWEJ STRAŻY POŻARNEJ, issued a contract notice for innovative high-performance computing infrastructure to support its Warning and Alarm System, explicitly linking AI applications to public safety.
Traditional research institutions are also scaling up. In November 2025, THE CYPRUS INSTITUTE launched a tender for high-performance AI servers to establish the Cyprus Edge-AI Hub for advanced research in artificial intelligence and related fields. In October 2025, Vysoká škola báňská - Technická univerzita Ostrava began work to modernise its equipment by acquiring a supercomputer, described as a comprehensive solution for high-performance computing and artificial intelligence that includes implementation, integration, training and maintenance services.
Major computing centres are looking beyond single projects. In October 2025, Garching, Leibniz-Rechenzentrum der Bayerischen Akademie der Wissenschaften issued a wide-ranging tender for the procurement of HPC and AI cluster components, alongside a public contract for a turnkey AI-optimised HPC solution. In November 2025, Vilniaus Gedimino technikos universitetas (PV) followed with a plan to expand a data processing and storage platform through four lots focused on CPU subsystems, GPU nodes, AI nodes and data storage infrastructure.
The pace did not slow towards the end of the year. On 17th November 2025, Rectorado de la Universidad Complutense de Madrid sought supercomputing equipment. And on 26th December 2025, the same day as the SURF B.V. notice, the Dirección General de la Fundación Computación y Tecnologías Avanzadas de Extremadura signalled plans to supply infrastructure and equipment for expanded data storage and high-performance GPU clusters, including installation and integration services.
Many 2025 procurements have focused on adding powerful servers to existing facilities. In July 2025, Mykolo Romerio universitetas (PV) launched a tender for a high-performance server cluster to support machine learning, artificial intelligence and high-performance computing services. In December 2025, Ústav organické chemie a biochemie AV ČR, v. v. i. sought AI GPU servers, including transport, installation and full operational setup. Universität Oldenburg’s AI server notice, published in September 2025, similarly combines delivery, installation and commissioning.
By contrast, a smaller group of buyers is procuring entire facilities. In July 2025, SDIS CANTAL issued a contract notice simply for the provision of a data center. Also in July 2025, Vojenské lesy a statky ČR, s.p. sought the supply of hardware for a data centre, explicitly including installation, technical support, warranty and proactive services. In September 2025, the Presidencia de la Diputación Provincial de Soria flagged its intention, via a prior information notice, to procure the supply of a Data Processing Center.
Seen against this backdrop, SURF B.V.’s move to acquire a complete high-performance computing datacentre for its AI facility fits a broader pattern. Rather than adding scattered capacity, public bodies are designing environments where compute, storage, networking, power and physical infrastructure can be tuned together to support demanding AI workloads. The Lithuanian and Czech higher education tenders, with their emphasis on integrated platforms and long-term maintenance, underline how AI is beginning to reshape not only applications but the underlying infrastructure strategies.
The SURF B.V. notice is succinct, and many practical questions remain open: the eventual scale of the facility, how it will be integrated with existing systems, and the service levels the AI facility will be expected to deliver. However, the direction is clear. Like their counterparts across Europe, the contracting authority is putting compute infrastructure at the heart of its approach to artificial intelligence.
As contract awards follow for this and related procurements, observers will be watching how far public organisations use these new datacentres to share capacity, standardise AI platforms and support a growing mix of research, operational and monitoring tasks. For now, the decision to commission an entire high-performance computing datacentre for AI is a strong signal that, in the public sector, the foundations of AI are increasingly being laid in concrete, cables and racks rather than just in software pilots.
Follow Tenderlake on LinkedIn for concise insights on public-sector tenders and emerging procurement signals.