believing before others understand: building europe's first trillion dollar company

Proxima Fusion founders Martin Kubie, Jonathan Schilling, Lucio Milanese, Jorrit Lion, and Francesco Sciortino

Recently, fusion silently moved from speculative science to an engineering race. redalpine exists for precisely these moments - when a fringe engineering breakthrough tips over into inevitability. 

In the summer of 2023, we noticed something unusual happening under the surface: critical supporting technologies had quietly crossed important thresholds simultaneously. High-temperature superconducting (HTS) tapes, once reserved for only the most exotic applications, began following the path solar panels took in the early 2010s - moving from lab curiosities to components engineers could finally afford to design around. 

Computational modeling, once gated by prohibitively expensive supercomputing time, now rides the cost-performance curve of NVIDIA’s GPUs - driven by the AI boom - enabling fusion-grade simulations to run on commodity hardware in a matter of hours. These shifts compress the risk stack from plasma physics to precision engineering - a discipline where Europe excels. Early validation is already in the data: Wendelstein 7-X demonstrated stable, disruption-free plasma, upon which Proxima’s team abstracted stellarator design into a repeatable optimisation problem; and the market has recognised the moment, responding with a record €130M Series A into the Munich-based team whose seed round we led just one year ago.

Often the most transformative changes happen quietly, noticed first by obsessives rather than headlines. Their enabling technologies accumulate quietly until reaching an inflection point. It was reminiscent of how ChatGPT appeared overnight, built atop decades-old graphics processors traced back to 1995 gaming technology, or how Tesla’s success hinged on lithium-ion batteries first designed for portable music players which were then scaled through decades of phones and laptops. Fusion hit that same quiet tipping point in 2023 and we recognized a moment others overlooked: fusion had suddenly become investable.

There are lots of ways people have tried to make fusion work - lasers, pinch devices, inertial tricks. Most of them sound clever, and some even look good in a lab. But only one category has really settled the core physics: magnetic confinement (MCF). Within that, there are two main architectures. Tokamaks rose to prominence in the 2010s, partly because their geometry seemed simple and their path to ignition more intuitive. But that simplicity masked a structural flaw: they rely on plasma current to maintain part of the magnetic field. That current introduces inherent instability. Stellarators took the harder route - mathematically and computationally - but in doing so, they eliminated the stability problem.

A visit to ITER

That clarity only deepened when we visited ITER, which is a staggering endeavor set up as a multinational collaboration involving 35 nations, spread across decades and continents. Under construction in southern France, it's easily the largest and most intricate fusion machine ever conceived: 400,000 tons of concrete and steel, 10,000 tons of superconducting magnets, and countless interconnected systems engineered to exquisite tolerances.

But what's striking isn't just ITER's ambition - it's the complexity born directly from the underlying physics. Tokamaks depend fundamentally on plasma currents, and those currents inevitably become unstable, producing sudden and violent disruptions. ITER isn't just big because it's bold; it's big because it has to be. Its size and redundancy - advanced feedback loops, correction coils, massive divertors, sacrificial armor tiles - are all responses to an instability at the very core of the design.

The stellarator flips this approach on its head and fundamentally reshapes fusion’s stability challenge. Rather than constantly battling disruptions, it eliminates their underlying cause - the plasma current itself. By creating magnetic fields purely through external coils, stellarators shift stability from a reactive, control-dependent struggle into an inherent structural property. The problem moves upstream, solved through meticulous geometry and advanced manufacturing. This approach exchanges active intervention for pre-engineered precision, swapping layers of real-time complexity for upfront intellectual rigor. 

Source: Caballero Tello, J. M. (2023). MSc thesis (Fig. 1.2). Eindhoven University of Technology.

Wendelstein 7-X was a critical proof of that concept. In 2018, it demonstrated that stellarators could genuinely hold stable plasma - no current, no disruptions, no elaborate real-time corrections. But it left an important question hanging: could these finely tuned magnetic geometries be systematically designed, or would each stellarator remain a unique masterpiece, requiring enormous manual effort?

In 2022, Landreman and Paul tackled exactly that problem. They showed you could actually optimize stellarator geometries computationally - even at realistic plasma pressures - by cleverly leveraging existing tokamak models. Their insight was subtle but powerful: quasi-symmetric stellarators behave neoclassically like axisymmetric tokamaks. That meant the fast, accurate bootstrap current calculations already developed for tokamaks could now be efficiently applied to stellarators. With this trick, they turned what was previously a nearly impossible computational burden into something straightforward enough for routine optimization.

Goodman and his collaborators pushed further in 2023. They developed new numerical methods specifically targeting quasi-isodynamic configurations - stellarators designed explicitly to minimize energy loss from fast particles and reduce neoclassical transport to negligible levels. By systematically refining the magnetic field, they found geometries that performed substantially better, even under realistic operating conditions. Crucially, their solutions weren't delicate artifacts. They proved robust, retaining their advantageous properties when finite plasma pressures were introduced.

Together, these two steps transformed stellarator design. Instead of trial-and-error, stellarator shapes could now be deliberately and repeatedly optimized. The physics moved from uncertain to predictable, and engineering became the primary challenge - hard, but fundamentally solvable. On our internal map of fusion risk, stellarators suddenly occupied a uniquely promising region: the physics was essentially done, and what remained was executing the engineering at scale.

One evening, Harald and I sketched out what the perfect stellarator startup might look like: simulation-first, speed-obsessed, and built around the clear strategic insight that the physics was solved, and the real frontier was fast, precise engineering. It would prioritize software over hardware, iterate in silico before touching metal, and move like a startup, not a research institute. No legacy constraints - just a clean sheet and a deadline. 

Not long after, we walked into Proxima Fusion’s first humble office in Munich - and found a team that had independently landed on almost the exact same answers. Quietly, Proxima had assembled something rare: a genuinely world-class team. Not just talented people, but exactly the right kind of talent - physicists and engineers who had trained at the leading labs globally, but then chosen to return to Europe. 

This wasn't by accident. Stellarators depend crucially on microscopic precision; the tiniest manufacturing imperfections translate directly into degraded performance. And precision, perhaps more than innovation or scale, has long been Europe's hidden advantage.

Europe's industrial ecosystem is dense with specialist firms built precisely to solve these high-stakes, high-precision challenges. Alongside this deep expertise, Proxima developed their own specialized computational engine, StarFinder. It turned the painstaking work of stellarator optimization into something that could be iterated in hours rather than months. With StarFinder churning out precise reactor geometries practically overnight, the critical bottleneck shifted from theory and computation to practical, careful execution. They had all the right ingredients - talent, technology, and technique - now it is all about carefully putting those pieces together. 

That’s when we pre-empted their seed round and simply told them: accelerate. Because in energy, the hard part isn’t scale - it’s crossing the threshold where something actually works. After that, things move fast.

In the 1970s, France didn’t ease into nuclear. They saw the fundamentals were sound and built as if time mattered. Within a decade, they went from almost nothing to 80% of their electricity from reactors. That’s how real energy transitions happen - not slowly, but all at once, when the physics stops being the bottleneck.

Fusion today sits poised to replicate that leap, not just in one country but globally. It aligns with the quiet tectonic shifts happening beneath the surface: exploding energy demands from data centers, AI workloads, and digital infrastructure that renewables alone simply can’t handle.

Proxima's Stellaris reactor - designed explicitly for gigawatt-scale - already projects energy costs comfortably below fossil fuels, even without pricing in climate externalities. Fusion has quietly slipped past that tipping point. The question is no longer "if" but "how quickly" the infrastructure comes online. Fusion doesn't just supplement our energy production - it redefines humanity's relationship with energy altogether, pushing civilization meaningfully up the Kardashev scale.

Just one year after we led their €20M Seed round with a pre-emptive offer, Proxima Fusion has taken another bold leap forward. Today, we proudly congratulate Proxima Fusion on raising their remarkable €130M Series A - the largest private fusion investment ever seen in Europe. This milestone, co-led by Cherry Ventures and Balderton Capital, with substantial backing from UVC Partners, DeepTech & Climate Fonds (DTCF), Plural, Leitmotif, Lightspeed, Bayern Kapital, HTGF, Club degli Investitori, Omnes Capital, Elaia Partners, Visionaries Tomorrow, Wilbe, and ourselves at redalpine, marks a definitive step toward the world's first stellarator-based fusion power plant. 

With over €185 million in total funding now secured, Proxima Fusion is decisively positioned to deliver commercial fusion power in the 2030s, strengthening Europe's energy resilience and setting a global benchmark in clean, limitless energy innovation.

This article was written by redalpine Investment Manager Gianmarco Hodel.