top of page
Search

Silicon Fjord: The New Rules of High-Tech Sovereignty in the Nordic Region

Futuristic data centers on snowy cliffs, glowing circuits in icy water, and Northern Lights in a digital landscape.

1. Introduction: The Architecture of Sovereignty

The mid-2020s have witnessed a profound transformation in the scientific posture of the Nordic nations. Sweden, Denmark, Finland, Norway, and Iceland—long celebrated for their social welfare models and environmental stewardship—have rapidly evolved into a cohesive bloc of "deep technology" innovation. This shift is not merely industrial; it is rooted in a fundamental reimaging of how scientific infrastructure interacts with the natural world and geopolitical necessity.

The driving force behind this renaissance is the concept of "technological sovereignty." As global supply chains fracture and great power competition intensifies, the Nordic Council of Ministers and national governments have recognized that reliance on external providers for critical digital and physical infrastructure is a strategic vulnerability.1 Consequently, the region has pursued a "triple helix" strategy, tightly weaving together state policy, academic research, and private enterprise to secure independent capabilities in four pivotal domains: Space, High-Performance Computing (HPC), Artificial Intelligence (AI), and Robotics.

This report provides an exhaustive scientific analysis of these breakthroughs. It moves beyond the headlines to explore the underlying physics, engineering challenges, and algorithmic innovations that define this new era. We examine how the thermodynamics of Finnish supercomputers are rewriting the rules of district heating, how the control theory of Swedish reusable rockets is enabling mainland European access to orbit, and how the quantum mechanics of Danish algorithms are accelerating the discovery of life-saving pharmaceuticals. Throughout this narrative, a distinct "Nordic signature" emerges: a relentless pursuit of performance that is inextricably linked to sustainability and resilience.

2. The Arctic Spaceport: Physics and Engineering at the Edge of Orbit

The Arctic has long been a theater for atmospheric research, but in the period from 2024 to 2026, it has cemented its status as Europe's primary mainland gateway to space. The focal point of this transformation is the Esrange Space Center in Kiruna, Sweden, which has graduated from suborbital sounding rocket campaigns to hosting the testing and launch of orbital-class reusable vehicles.3

2.1 The Themis Demonstrator and the Physics of Reusability

The most significant engineering endeavor in Nordic spaceflight is the Themis reusable booster program. Initiated by the European Space Agency (ESA) and led by ArianeGroup, Themis represents Europe's concerted effort to master the complex physics of Vertical Takeoff and Vertical Landing (VTOL) rocketry—a capability previously dominated by American entities.4

2.1.1 Aerodynamic Control and the "Hop" Tests

In 2025, the Themis demonstrator, a 30-meter-tall stainless steel vehicle, was positioned at Esrange for a series of "hop tests." These tests are critical for validating the flight control algorithms required for reusability. A returning rocket stage behaves fundamentally differently from a standard projectile; during descent, it acts as an inverted pendulum balanced on a column of thrust.

The control system must manage high-frequency perturbations caused by atmospheric turbulence. To maintain stability, Themis employs grid fins—latticed aerodynamic control surfaces that remain effective across a wide range of Mach numbers, from hypersonic reentry to subsonic landing. The "hop" profile involves lifting off to a low altitude, hovering, and then executing a precision landing. This maneuver demands millisecond-level synchronization between the engine's thrust vector control (TVC) and the reaction control system (RCS). The choice of Esrange for these tests is dictated by safety physics; the vast, unpopulated impact corridors of the Swedish Lapland provide the necessary risk mitigation for experimental guidance, navigation, and control (GNC) architectures that would be too hazardous to test in densely populated Central Europe.6

Although meteorological conditions—specifically heavy snow accumulation in Kiruna—delayed the initial flight campaigns until the spring thaw of 2026, the static fire and wet dress rehearsals successfully validated the cryogenic fluid handling systems in extreme cold.9

2.2 The Prometheus Engine: Methalox Propulsion Cycles

At the heart of Themis lies the Prometheus engine, a technological breakthrough that shifts European propulsion from the hydrogen-oxygen cycles of the Ariane 5 to a methane-oxygen (methalox) cycle. This transition is driven by specific thermochemical and operational advantages.10

2.2.1 Thermodynamics of Methane vs. Hydrogen

Prometheus burns Liquid Natural Gas (LNG/Methane) and Liquid Oxygen (LOx). The scientific rationale for this switch is threefold:

  1. Density Impulse: Liquid methane is approximately six times denser than liquid hydrogen (LH2). While hydrogen offers a higher specific impulse (efficiency by weight), its low density requires enormous fuel tanks, increasing the structural mass and drag of the vehicle. Methane allows for compact tankage, which is critical for a reusable stage that must withstand the mechanical stress of reentry and landing.12

  2. Thermal Compatibility: Methane and oxygen have similar boiling points (-161°C and -183°C, respectively). This allows for "common bulkhead" tank designs, where the fuel and oxidizer share a structural wall, reducing weight. In contrast, hydrogen (-253°C) requires heavy insulation to prevent it from freezing the oxygen or boiling off.10

  3. Coking and Reusability: Kerosene engines suffer from "coking"—the deposition of carbon soot in the cooling channels and injectors, which requires extensive cleaning between flights. Methane burns cleanly, leaving negligible residue. This chemical property is the lynchpin of rapid reusability, minimizing refurbishment time.10

Furthermore, Prometheus extensively utilizes additive manufacturing (3D printing) for its combustion chamber and gas generator. This manufacturing technique reduces the component count by a factor of ten compared to the Vulcain engine, significantly lowering the cost of production and allowing for rapid iterative prototyping.12

2.3 Suborbital Science: Microgravity and Atmospheric Chemistry

While orbital launchers garner the spotlight, the scientific utility of sounding rockets at Esrange has reached new levels of sophistication. These vehicles provide a unique microgravity environment—approximately six minutes of high-quality weightlessness—free from the "g-jitter" (vibrations) present on crewed stations like the ISS.14

2.3.1 MAPHEUS-16 and the MOSAIC Platform

The MAPHEUS-16 mission, launched in November 2025, set a record by carrying 21 distinct experiments to an altitude of 267 kilometers. A key innovation on this flight was the MOSAIC (Micro-Experiments on Sounding Rockets As Insert Cubes) platform. This modular system standardizes payload integration using 10x10x10 cm cubes, democratizing access to space for smaller research teams.15

Scientific Experiments:

  • Solidification Physics: One primary experiment investigated the solidification of metallic alloys in microgravity. On Earth, gravity-driven convection currents disturb the melt as it cools, creating non-uniform grain structures. In the microgravity of the MAPHEUS flight, researchers successfully printed metal samples with perfectly uniform crystalline structures. This data is foundational for future "in-space manufacturing" (ISM), where large structures could be fabricated in orbit with material properties superior to those made on Earth.15

  • Cellular Biology (FLUMIAS): The mission also carried the FLUMIAS microscope, capable of high-resolution 3D fluorescence imaging of living cells. Researchers observed the cytoskeleton of human immune and cancer cells in real-time. The cytoskeleton is known to degrade in microgravity, leading to immune suppression in astronauts. By capturing the immediate structural collapse of these cellular scaffolds during the launch's hypergravity and subsequent microgravity phases, scientists gained crucial insights into the mechanobiology of spaceflight adaptation.17

2.3.2 ORIGIN-2: Probing the Mesosphere

In late 2025, the ORIGIN-2 sounding rocket mission explored the "nightglow" phenomenon—a faint luminescence in the upper atmosphere (80–100 km). This emission is caused by the recombination of atomic oxygen, which is created during the day when solar UV radiation splits oxygen molecules.18

Atmospheric Coupling:

The distribution of atomic oxygen is a critical variable in climate models. It acts as a chemical energy reservoir, influencing the thermal balance of the mesosphere and lower thermosphere (MLT) region. The ORIGIN-2 mission used advanced photometers to measure these emissions in situ. The data provides a "ground truth" for calibrating satellite instruments and improves the accuracy of models that couple the upper atmosphere with lower atmospheric weather patterns. Understanding this coupling is increasingly important as climate change alters the density and temperature profiles of the upper atmosphere, potentially affecting satellite drag and orbital lifetimes.18

3. High-Performance Computing: The Green Engine of Discovery

In the domain of computing, the Nordic region has leveraged its sub-Arctic climate and abundant renewable energy to build the world's most sustainable High-Performance Computing (HPC) ecosystem. The crown jewel of this effort is the LUMI supercomputer in Kajaani, Finland.

3.1 LUMI: Architecture of the Pre-Exascale Era

Owned by the EuroHPC Joint Undertaking and a consortium of ten countries, LUMI is a pre-exascale system with a sustained performance exceeding 380 petaflops and a peak performance over 550 petaflops. Its architecture is explicitly designed for the convergence of simulation and Artificial Intelligence.20

3.1.1 GPU Acceleration and the MI250X

LUMI's computational power is derived from its "LUMI-G" partition, which houses thousands of AMD Instinct MI250X Graphics Processing Units (GPUs). Unlike traditional Central Processing Units (CPUs) which process tasks serially, GPUs are massively parallel, making them ideal for grid-based physical simulations and matrix-based AI training.22

Scientific Application: Solar Dynamos (VISSI)

In 2025, the VISSI project utilized LUMI's GPU partition to solve one of astrophysics' enduring mysteries: the solar dynamo. Previous models could not replicate the small-scale magnetic fluctuations in the sun's convection zone that eventually coalesce into sunspots and flares. LUMI's massive parallelism allowed researchers to resolve these small-scale fluid dynamics for the first time. The resulting high-fidelity simulations confirmed the existence of a "fluctuating dynamo" mechanism, significantly improving the predictive models for space weather events that can disrupt power grids and telecommunications on Earth.23

3.2 Thermodynamics of Computation: Waste Heat Recovery

LUMI's most revolutionary feature is not its speed, but its integration into the local energy grid. Large supercomputers generate immense heat; traditionally, this energy is vented into the atmosphere using evaporative cooling towers, consuming vast amounts of water and electricity.

3.2.1 District Heating Integration

LUMI employs a direct liquid cooling system where water circulates directly over the processor dies, capturing heat at a high efficiency. This heated water (approx. 30–40°C) is then pumped into heat pumps which boost the temperature further, making it suitable for the district heating network of Kajaani.

In 2025, this system supplied approximately 20% of the city's annual heating demand. This "negative carbon" footprint—where the computer effectively replaces the burning of biomass or fossil fuels for heat—reduces the city's CO2 emissions by an estimated 12,000 tons annually. This model establishes a new standard for HPC facility engineering, proving that digital infrastructure can act as a thermal battery for municipal systems.24

3.3 Destination Earth: The Climate Change Adaptation Digital Twin

LUMI serves as the primary host for the Climate Change Adaptation Digital Twin (Climate DT), a cornerstone of the European Commission's Destination Earth (DestinE) initiative. This project represents a fundamental shift in climate science: the move from statistical approximations to deterministic, kilometer-scale resolving of Earth systems.27

3.3.1 Resolving Convection at Kilometer Scale

Traditional Global Circulation Models (GCMs) operate at resolutions of 50–100 km. At this scale, crucial phenomena like convective storm systems, ocean eddies, and local topography must be "parameterized"—approximated using statistical rules. Parameterization is a major source of uncertainty in climate projections.

The Climate DT running on LUMI achieves a global resolution of 5–10 km. At this scale, the physics of deep convection (thunderstorms) are explicitly resolved rather than approximated. This allows for the realistic simulation of extreme precipitation events and wind storms that traditional models often miss or underestimate.29

3.3.2 "Storyline" Simulations and Policy Impact

A key innovation of the Climate DT is the ability to run "storyline" simulations—interactive, on-demand scenarios. Following the catastrophic Storm Boris in Central Europe in 2024, researchers used the Climate DT to simulate the same weather event under different future warming scenarios (e.g., +2°C world vs. +4°C world).

These simulations provided policymakers with concrete, localized data on flood depths and wind speeds for specific cities, moving the discourse from abstract global averages to tangible infrastructure risks. The system has also been used to model sea-ice drift in the Gulf of Riga to support offshore wind farm planning, demonstrating its direct utility for the green energy transition.29

4. Quantum Technologies: The Cold Logic of the North

While HPC scales up classical physics, Nordic researchers are making fundamental breakthroughs in quantum information science. The region's strategy focuses on two complementary axes: building robust quantum hardware (Sweden) and developing specialized quantum algorithms for life sciences (Denmark).

4.1 WACQT and the Challenge of Connectivity

The Wallenberg Centre for Quantum Technology (WACQT), based at Chalmers University of Technology in Sweden, is spearheading the development of superconducting quantum computers. The program's roadmap targets a 100-qubit processor, a threshold where quantum systems begin to offer "quantum advantage" over classical supercomputers.33

4.1.1 The Flip-Chip Architecture Breakthrough

Scaling superconducting qubits is limited by a geometric problem: "crowding." Each qubit requires control lines (wires) to manipulate its state and resonators to read it out. On a 2D chip, routing these wires to the center of a large qubit array becomes impossible without interfering with the qubits themselves.

To solve this, WACQT researchers have successfully implemented a flip-chip (3D integrated) architecture. This design separates the system into two stacked layers:

  1. The Q-Chip: Containing the sensitive superconducting transmon qubits.

  2. The C-Chip: Containing the control wiring, readout resonators, and interconnects.The two chips are bonded together using superconducting indium bumps. In 2025, the team demonstrated that they could transmit signals across this interface while suppressing signal crosstalk—the unwanted electromagnetic interference between channels—to below -27 dB. This level of isolation is a critical engineering milestone, as it ensures that controlling one qubit does not accidentally corrupt the state of its neighbor, a prerequisite for error-corrected quantum computing.34

4.1.2 Kinetic Inductance Parametric Amplifiers (TWPA)

Another critical component developed in the Nordics is the Traveling Wave Parametric Amplifier (TWPA). Reading out a qubit involves detecting a microwave signal as weak as a single photon. Conventional amplifiers introduce too much thermal noise, drowning out the quantum signal.

Nordic researchers have perfected TWPAs using Kinetic Inductance. In superconducting materials like Niobium Titanium Nitride (NbTiN), the inductance (resistance to changing current) varies non-linearly with the current intensity. By pumping this material with a strong reference tone, the non-linearity drives a "four-wave mixing" process that transfers energy from the pump to the weak signal. This provides amplification with added noise near the fundamental quantum limit, allowing for high-fidelity single-shot readout of qubits.37

4.2 Kvantify: Quantum Chemistry and Drug Discovery

While Sweden builds the engine, Denmark is writing the navigation software. Kvantify, a spin-out from the Niels Bohr Institute, has emerged as a leader in quantum algorithms for chemistry.40

4.2.1 The FAST-VQE Algorithm

Simulating molecular bonds is one of the most promising applications of quantum computing. The standard algorithm for this is the Variational Quantum Eigensolver (VQE). However, VQE suffers from the "barren plateau" problem—a situation where the optimization landscape becomes flat, and the algorithm gets stuck, unable to find the lowest energy state of the molecule.

Kvantify's breakthrough is the FAST-VQE algorithm, which optimizes the initialization of the quantum state to start closer to the solution, avoiding these plateaus. In 2025, Kvantify demonstrated the ability to run this algorithm on a 50-qubit processor (IQM Emerald), calculating the binding free energy of drug candidates. Binding free energy determines how strongly a drug molecule attaches to its target protein; accurate calculation is the "holy grail" of pharmaceutical research. By achieving chemical accuracy on noisy intermediate-scale quantum (NISQ) hardware, Kvantify has bridged the gap between theoretical quantum advantage and practical industrial application.41

5. Artificial Intelligence: Sovereign Models and Grid Physics

The Nordic approach to AI diverges from the general-purpose, black-box models of Silicon Valley. Instead, the focus is on "Sovereign AI"—models that are transparent, linguistically inclusive, and integrated into critical infrastructure.

5.1 The Viking and Poro Model Families

The dominance of English-centric Large Language Models (LLMs) poses a cultural threat to smaller language groups. In response, Silo AI (Finland) and the University of Turku developed the Poro and Viking open-source models.44

5.1.1 Architectural Innovations for Low-Resource Languages

The Viking models (up to 33 billion parameters) are trained natively on Nordic languages (Finnish, Swedish, Danish, Norwegian, Icelandic) alongside English and code. Training an LLM on Finnish—an agglutinative language where a single word can have thousands of forms—requires specific architectural choices.

Viking utilizes Flash Attention and Rotary Positional Embeddings (RoPE).

  • Flash Attention: This memory-optimization technique allows the model to handle very long context windows (4096 tokens and beyond) efficiently. This is crucial for processing lengthy government documents or legal statutes, which are primary use cases for sovereign AI.

  • Rotary Embeddings: This mathematical encoding of token position improves the model's ability to understand relative relationships between words, which is vital for the complex syntax of Nordic languages.By training these models on the LUMI supercomputer, the project ensures that the model weights—and the cultural knowledge they encode—remain under European jurisdiction, fulfilling the mandate of digital sovereignty.22

5.2 AI as a Grid Stabilizer: The Frequency Containment Reserve

A novel intersection of AI and energy engineering is the VTT & RISE project, which treats data centers not just as consumers of power, but as active components of the electrical grid.46

5.2.1 The Physics of 50 Hz

The AC power grid must maintain a frequency of exactly 50 Hz. Deviations occur when supply and demand are mismatched. Traditionally, heavy flywheels or gas turbines provide the inertia to stabilize the grid.

The VTT/RISE pilot demonstrated that an HPC center could provide Frequency Containment Reserve (FCR). By using AI to monitor grid frequency and instantly throttling the clock speeds of thousands of GPUs, the data center can shed megawatts of load in milliseconds. This rapid response acts as a "virtual battery," stabilizing the grid during sudden drops in wind or solar generation. This breakthrough turns the energy-intensive AI industry into a solution for the renewable energy transition, providing a scalable mechanism for demand-side response.47

6. Robotics: Automating the Elements

Nordic robotics is defined by its environment: the stormy North Sea, the complex archipelago coastlines, and the high-wage industrial economy. These factors drive a focus on autonomous maritime operations and collaborative industrial robotics.

6.1 Reach Remote: Uncrewed Maritime Operations

Norway's Reach Subsea, in partnership with Kongsberg Maritime, has launched the Reach Remote USV (Uncrewed Surface Vessel) fleet. Delivered in 2025, these 24-meter vessels operate without a crew for up to 30 days, performing subsea inspection and intervention tasks.49

6.1.1 Active Heave Compensation and Tether Management

The critical engineering challenge for Reach Remote is deploying a Remotely Operated Vehicle (ROV) from a small, unmanned vessel in rough seas. On a large crewed ship, the vessel's mass dampens the motion of the waves. On a light USV, the vessel pitches and rolls violently.

To prevent the ROV from being yanked by its tether, the USV employs an advanced Active Heave Compensation (AHC) system. This system uses motion reference units (accelerometers and gyroscopes) to predict the vessel's movement and rapidly pays out or reels in the tether to keep the ROV stationary relative to the seabed. This robotic tether management allows the system to perform delicate valve manipulations on the seafloor while the surface vessel is buffeted by North Sea waves. This capability allows the USV to replace crewed vessels that are 50 times larger, reducing mission emissions by over 90%.51

6.2 Zeabuz and Torghatten: Sensor Fusion on the Water

In urban maritime transport, the MF Estelle in Stockholm and upcoming hydrogen ferries in Norway are pioneering autonomous navigation using technology from the spin-off Zeabuz.53

6.2.1 Multi-Modal Sensor Fusion

Navigating a busy harbor requires a level of situational awareness that exceeds standard radar. The Zeabuz system employs a sensor fusion engine that integrates data from:

  • Lidar: For precise, centimeter-level 3D mapping of docks and nearby obstacles.

  • Computer Vision (Optical/Thermal): For object classification (distinguishing a kayak from a swan or a buoy).

  • Radar: For speed and trajectory estimation of larger vessels.An AI "Captain" synthesizes these streams to generate a dynamic collision avoidance path. Unlike open-ocean autopilots, this system must understand the "rules of the road" in constrained waterways. In 2025, the system moved towards "Remote Operation Center" (ROC) supervision, where a single human captain on shore monitors multiple ferries, stepping in only when the AI encounters an ambiguous situation.54

6.3 Universal Robots: The AI Accelerator

Denmark's Universal Robots (UR) continues to lead the "cobot" (collaborative robot) sector. The major advancement in 2024–2025 is the AI Accelerator hardware toolkit, which embeds NVIDIA Jetson edge AI computing directly into the robot controller.55

6.3.2 Dynamic Path Planning vs. Fixed Waypoints

Traditional industrial robots follow fixed spatial coordinates. If an obstacle enters their path, they must emergency stop. The AI Accelerator enables Dynamic Path Planning. Using attached 3D cameras, the robot builds a voxel map of its environment in real-time (milliseconds). If a human arm or a misplaced bin enters its trajectory, the robot's motion planner recalculates a new, collision-free curve to the target without stopping.

This capability is essential for "chaotic bin picking"—grasping randomly oriented parts from a bin. The AI uses pose estimation to identify the best grip point on a part and plans a path that avoids collisions with the bin walls or other parts. This innovation lowers the barrier to entry for automation in high-mix, low-volume manufacturing, a staple of the European SME landscape.57

7. Conclusion: The Integrated Future

The scientific landscape of the Nordic countries in the mid-2020s serves as a blueprint for the future of high-technology infrastructure. The region has moved beyond the passive consumption of digital services to become an active architect of the hardware and software that underpins them.

From the Prometheus engines lighting up the Arctic sky with clean methane flames 10 to the LUMI supercomputer warming homes with the waste heat of climate simulations 24, the "Nordic signature" is clear: high performance is inseparable from environmental sustainability. The successful testing of the Themis reusable booster 8 and the deployment of the Reach Remote autonomous fleet 50 demonstrate a mastery of complex control theory in harsh physical environments. Meanwhile, the Viking AI models 22 and WACQT quantum processors 34 ensure that the region retains sovereign control over the intellectual and computational substrates of the future.

As these technologies mature, their convergence offers even greater potential. We can anticipate Quantum algorithms optimizing the AI path planning of autonomous vessels, all coordinated via space-based assets launched from Esrange. In doing so, the Nordic nations are not merely observing the technological horizon; they are defining it.

Table 1: Key Nordic Scientific Breakthroughs & Infrastructure (2024–2026)

Domain

Project/Facility

Country

Key Technical Innovation

Scientific/Industrial Impact

Space

Themis & Prometheus

Sweden (EU)

Methalox propellant cycle; VTOL control algorithms.

Sovereign reusable launch capability; Green propulsion.

Space

MAPHEUS-16

Sweden (DLR)

MOSAIC modular payload; Microgravity metal solidification.

In-space manufacturing; Biological research.

HPC

LUMI

Finland

Waste heat recovery (district heating); GPU-based physics.

Climate Digital Twins; Sustainable Exascale computing.

HPC

Climate DT

Finland (EU)

Kilometer-scale resolution; Explicit convection solving.

Accurate local climate adaptation planning.

Quantum

WACQT Processor

Sweden

Flip-chip 3D integration; -27dB crosstalk suppression.

Scalable superconducting quantum hardware.

Quantum

Kvantify FAST-VQE

Denmark

Initialization optimization for chemistry algorithms.

Drug discovery (binding free energy calculation).

AI

Viking / Poro

Finland

Flash Attention; Rotary Embeddings for Nordic languages.

Sovereign, linguistically inclusive LLMs.

AI/Energy

VTT/RISE FCR

Fin/Swe

GPU throttling for grid frequency containment.

Grid stability; Demand response integration.

Robotics

Reach Remote

Norway

Active Heave Compensation for tethered ROVs.

Unmanned offshore operations; 90% emission cut.

Robotics

Zeabuz

Norway/Swe

Multi-modal sensor fusion (Lidar/Radar/Vision).

Autonomous urban ferries.

Robotics

UR AI Accelerator

Denmark

Dynamic path planning via NVIDIA Jetson edge AI.

Chaotic bin picking; Adaptive industrial automation.

Table 2: Technical Comparison of Nordic Compute Architectures

Feature

LUMI Supercomputer (HPC)

WACQT Quantum Processor

Core Unit

AMD Instinct MI250X GPU

Superconducting Transmon Qubit

Scale

>550 Petaflops (Classical)

25 Qubits (Quantum)

Cooling

Direct Liquid Cooling (Water)

Dilution Refrigerator (<20 mK)

Waste Heat

Reused for District Heating

Negligible (Dissipated)

Primary Physics

Classical Fluid Dynamics / Matrix Math

Quantum Superposition / Entanglement

Key Use Case

Climate Modeling, LLM Training

Quantum Chemistry, Optimization

Works cited

  1. Digital Sovereignty in Europe – a first benchmark, accessed January 9, 2026, https://www.wik.org/fileadmin/files/_migrated/news_files/Digital_Sovereignty_Report.pdf

  2. European Software and Cyber Dependencies, accessed January 9, 2026, https://www.europarl.europa.eu/RegData/etudes/STUD/2025/778576/ECTI_STU(2025)778576_EN.pdf

  3. Orbital launches from Esrange Space Center: - Luftfartstilsynet, accessed January 9, 2026, https://www.luftfartstilsynet.no/globalassets/dokumenter/romfart/report_orbital-launches-from-esrange-space-center.pdf

  4. Top European Launch Companies of 2025 - European Spaceflight, accessed January 9, 2026, https://europeanspaceflight.com/top-european-launch-companies-of-2025/

  5. Themis demonstrator ready for final tests at ESRANGE, accessed January 9, 2026, https://hadea.ec.europa.eu/news/themis-demonstrator-ready-final-tests-esrange-2025-09-19_en

  6. Themis installed on its launch pad - Swedish Space Corporation, accessed January 9, 2026, https://sscspace.com/themis-stands-on-the-launch-pad/

  7. Themis programme - Wikipedia, accessed January 9, 2026, https://en.wikipedia.org/wiki/Themis_programme

  8. Europe's first steps to master rocket recovery and re-use – happening soon at Esrange, accessed January 9, 2026, https://sscspace.com/europes-first-steps-to-master-rocket-recovery-and-reuse-happening-soon-at-esrange/

  9. ESA “Waiting for Spring” to Launch Themis Reusable Rocket Demonstrator, accessed January 9, 2026, https://europeanspaceflight.com/esa-waiting-for-spring-to-launch-themis-reusable-rocket-demonstrator/

  10. ESA - Prometheus - European Space Agency, accessed January 9, 2026, https://www.esa.int/Enabling_Support/Space_Transportation/Future_space_transportation/Prometheus

  11. ESA - Themis - European Space Agency, accessed January 9, 2026, https://www.esa.int/Enabling_Support/Space_Transportation/Themis

  12. PROMETHEUS | CNES, accessed January 9, 2026, https://cnes.fr/en/projects/prometheus

  13. PROMETHEUS, A LOX/LCH4 REUSABLE ROCKET ENGINE - eucass, accessed January 9, 2026, https://www.eucass.eu/doi/EUCASS2017-537.pdf

  14. Sounding rocket launched from ESRANGE - | SPACECONOMY, accessed January 9, 2026, https://spaceconomy.fi/mapheus16

  15. MAPHEUS-16 launched: Sets new record - SSC - Swedish Space Corporation, accessed January 9, 2026, https://sscspace.com/mapheus-16-launched/

  16. MAPHEUS-16, accessed January 9, 2026, https://www.dlr.de/en/so/research-transfer/research-infrastructure/sounding-rockets/mapheus-16

  17. 24th ESA Symposium on European - ROCKET & BALLOON - programmes and related research - TIB, accessed January 9, 2026, https://edocs.tib.eu/files/e01mr20/1687597367.pdf

  18. Successful launch of sounding rocket for MISU project ORIGIN to study nightglow in Earth's atmosphere - Stockholms universitet, accessed January 9, 2026, https://www.su.se/english/divisions/department-of-meteorology/news/articles/2025-10-30-successful-launch-of-sounding-rocket-for-misu-project-origin-to-study-nightglow-in-earths-atmosphere

  19. Research rocket ORIGIN-2 studying nightglow phenomenon - SSC - Swedish Space Corporation, accessed January 9, 2026, https://sscspace.com/origin2/

  20. LUMI, Europe's most powerful supercomputer, is solving global challenges and promoting a green transformation, accessed January 9, 2026, https://www.vsb.cz/veda/en/news-detail/?reportId=43792&linkBack=%2Fveda%2Fen%2Findex.html

  21. Sustainability in HPC, a look at LUMI Fredrik Robertsén, accessed January 9, 2026, https://www.bnl.gov/modsim/events/2024/files/talks/fredrik-robertsen.pdf

  22. Viking 13B: Scaling Nordic AI Models Using an Open Source Training Framework for LUMI, accessed January 9, 2026, https://www.amd.com/en/blogs/2024/viking-13b-scaling-nordic-ai-models-using-an-open-source-training-framework-for-lumi.html

  23. Study Utilizing LUMI Supercomputer Offers New Perspectives on Solar Storm Predictions, accessed January 9, 2026, https://www.hpcwire.com/off-the-wire/study-utilizing-lumi-supercomputer-offers-new-perspectives-on-solar-storm-predictions/

  24. LUMI, an HPE Cray EX system, has a sustained computing power of 380 petaflops (HPL, High-Performance Linpack) in its final configuration. LUMI is also one of the world's leading platforms for artificial intelligence. - LUMI supercomputer, accessed January 9, 2026, https://lumi-supercomputer.eu/lumi_supercomputer/

  25. Sustainability | LUMI AI Factory, accessed January 9, 2026, https://lumi-ai-factory.eu/about-us/sustainability/

  26. A supercomputer to heat houses in northern Finland - Granlund, accessed January 9, 2026, https://www.granlundgroup.com/news/a-supercomputer-to-heat-houses-in-northern-finland/

  27. DestinE 2025: European collaboration powering Digital Twins and AI, accessed January 9, 2026, https://destine.ecmwf.int/news/destine-2025-european-collaboration-powering-digital-twins-and-ai/

  28. Climate Change Adaptation Digital Twin - Destination Earth - ECMWF, accessed January 9, 2026, https://destine.ecmwf.int/climate-change-adaptation-digital-twin-climate-dt/

  29. Updates on the climate change adaptation Digital Twin - Destination Earth, accessed January 9, 2026, https://destination-earth.eu/news/updates-on-the-climate-change-adaptation-digital-twins/

  30. The Destination Earth digital twin for climate change adaptation - EGUsphere, accessed January 9, 2026, https://egusphere.copernicus.org/preprints/2025/egusphere-2025-2198/

  31. The Nordic Cryosphere Digital Twin (NOCOS DT) - Publications, accessed January 9, 2026, https://pub.norden.org/temanord2025-565/temanord2025-565.pdf

  32. Third Destination Earth User eXchange showcases the fast development of the system, accessed January 9, 2026, https://destine.ecmwf.int/news/third-destination-earth-user-exchange-showcases-the-fast-development/

  33. The initiative driving Sweden's quantum revolution | Knut and Alice Wallenberg Foundation, accessed January 9, 2026, https://kaw.wallenberg.org/en/research/initiative-driving-swedens-quantum-revolution

  34. Signal crosstalk in a flip-chip quantum processor - arXiv, accessed January 9, 2026, https://arxiv.org/html/2403.00285v2

  35. Signal Crosstalk in a Flip-Chip Quantum Processor - VTT's Research Information Portal, accessed January 9, 2026, https://cris.vtt.fi/en/publications/signal-crosstalk-in-a-flip-chip-quantum-processor/

  36. Signal Crosstalk in A Flip-Chip Quantum Processor | PDF - Scribd, accessed January 9, 2026, https://www.scribd.com/document/832110442/Signal-Crosstalk-in-a-Flip-Chip-Quantum-Processor

  37. A Traveling Wave Parametric Amplifier Isolator - arXiv, accessed January 9, 2026, https://arxiv.org/html/2406.19752v1

  38. A 4–8 GHz kinetic inductance traveling-wave parametric amplifier using four-wave mixing with near quantum-limited noise performance - AIP Publishing, accessed January 9, 2026, https://pubs.aip.org/aip/apq/article/1/3/036107/3302993/A-4-8-GHz-kinetic-inductance-traveling-wave

  39. Performance of a Kinetic Inductance Traveling-Wave Parametric Amplifier at 4 Kelvin: Toward an Alternative to Semiconductor Amplifiers - PMC - PubMed Central, accessed January 9, 2026, https://pmc.ncbi.nlm.nih.gov/articles/PMC10644704/

  40. Report: The New Nordic Innovation Ecosystem, accessed January 9, 2026, https://www.jetro.go.jp/ext_images/uk/NewNordic.pdf

  41. Kvantify Will Use Gefion's AI Supercomputer For Quantum Chemistry And Drug Discovery, accessed January 9, 2026, https://thequantuminsider.com/2025/05/13/kvantify-will-use-gefions-ai-supercomputer-for-quantum-chemistry-and-drug-discovery/

  42. Quantum Computing for Accurate Binding Free Energy in Drug Discovery - Kvantify, accessed January 9, 2026, https://www.kvantify.com/blog/quantum-computing-for-accurate-binding-free-energy-in-drug-discovery

  43. Kvantify is Scaling Quantum Chemistry on IQM Resonance, accessed January 9, 2026, https://meetiqm.com/blog/kvantify-is-scaling-quantum-chemistry-on-iqm/

  44. Read the 4th Nordic State of AI report - My AI, accessed January 9, 2026, https://my.ai.se/resources/5508

  45. Viking 33B: Scaling Performance of Open Nordic AI Models - AMD, accessed January 9, 2026, https://www.amd.com/en/blogs/2024/viking-33b-scaling-performance-of-open-nordic-ai-models.html

  46. Forging the Future Together: 20 Joint Nordic & International Innovation Projects Announced in 2025 - forumNordic, accessed January 9, 2026, https://forumnordic.com/business/forging-the-future-together-20-joint-nordic-international-innovation-projects-announced-in-2025/

  47. Project on Frequency services for the electrical grid from HPC data centers granted by Vinnova | RISE, accessed January 9, 2026, https://www.ri.se/en/project-on-frequency-services-for-the-electrical-grid-from-hpc-data-centers-granted-by-vinnova

  48. Sustainable Grid through Distributed Data Centers Spinning AI Demand for Grid Stabilization and Optimization© 2025 IEEE. Personal use of this material is permitted. Permission from IEEE must be obtained for all other uses, in any current or future media, including reprinting/republishing this material for advertising or promotional purposes, creating new collective works, for resale - arXiv, accessed January 9, 2026, https://arxiv.org/html/2504.03663v1

  49. Reach Subsea ASA orders Reach Remote 3 and 4 from Kongsberg Maritime, accessed January 9, 2026, https://reachsubsea.no/reach-subsea-asa-orders-reach-remote-3-and-4-from-kongsberg-maritime/

  50. Delivers first REACH REMOTE USV - KONGSBERG - an international technology group, accessed January 9, 2026, https://www.kongsberg.com/newsroom/news-archive/2025/delivers-first-reach-remote-usv/

  51. Reach Remote 1 & 2, accessed January 9, 2026, https://reachsubsea.no/assets/reach-remote-1-2/

  52. Reach Subsea confirms delivery of Reach Remote 1, accessed January 9, 2026, https://reachsubsea.no/reach-subsea-confirms-delivery-of-reach-remote-1/

  53. Torghatten and Zeabuz Make History in Stockholm, accessed January 9, 2026, https://www.torghatten.no/en/about/news/16b10b48-210f-4354-a6d9-d309e1aa3b34__i18n_en/torghatten-and-zeabuz-make-history-in-stockholm-launching-the-world-s-first-commercial-emission-free-autonomous-passenger-ferry-in-the-heart-of-the-city

  54. Torghatten's autonomous ferry key to combating congestion - Riviera Maritime Media, accessed January 9, 2026, https://www.rivieramm.com/news-content-hub/news-content-hub/torghattens-autonomous-ferry-key-to-combating-congestion-75440

  55. Universal Robots and Mobile Industrial Robots to Debut New and AI-powered Automation Solutions Across Integrated Industry Workflows at Automate 2025, accessed January 9, 2026, https://www.universal-robots.com/news-and-media/news-center/universal-robots-and-mobile-industrial-robots-to-debut-new-and-ai-powered-automation-solutions-across-integrated-industry-workflows-at-automate-2025/

  56. Universal Robots Launches Automated Robot Development with AI Accelerator - News, accessed January 9, 2026, https://control.com/news/universal-robots-launches-automated-robot-development-with-ai-accelerator/

  57. Universal Robots Offers AI Robotic Platforms at Automate 2025, accessed January 9, 2026, https://www.powertransmission.com/universal-robots-offers-ai-robotic-platforms-at-automate-2025

  58. Universal Robots Shows Deep Learning-Based Part Detection, accessed January 9, 2026, https://www.universal-robots.com/news-and-media/news-center/universal-robots-shows-deep-learning-based-part-detection-for-machine-tending-cobots-at-imts/

Comments

bottom of page