Fusion, Magnets, and the Race to Bottle a Star
Stephen Horton | February 2026
This is the companion blog post to The Silver Standard — a longer piece on fusion supply chains, geopolitics, and the silver market. This version focuses on the physics, the people, and the sheer audacity of trying to build a star on Earth.
The Professor, the Students, and the Tape That Changed Everything
In 2012, an MIT professor named Dennis Whyte gave his graduate students a homework assignment that would accidentally reshape the global fusion race.
The assignment was simple: design a fusion reactor using a new kind of superconducting magnet tape that had recently become commercially available. The tape — called REBCO (Rare Earth Barium Copper Oxide) — could carry enormous electrical currents with zero resistance, and it could do it at temperatures far warmer than the old superconductors that had been the industry standard for decades.
The students ran the numbers. Then they ran them again. What came back was startling.
With REBCO magnets, you could build a fusion device that was dramatically smaller than anything anyone had designed before — but that produced the same amount of fusion power. The magnets were so much stronger that you could squeeze the same physics into a fraction of the space. It was like discovering that the 747-sized machine everyone had been trying to build could actually fit in a garage.
Whyte, his postdoc, and three of those students didn’t just write a paper about it. They started a company: Commonwealth Fusion Systems (CFS). And they set out to outrun the largest science project on Earth.
David vs. Goliath: CFS vs. ITER
To understand what CFS did, you need to understand what they were racing against.
ITER — the International Thermonuclear Experimental Reactor — is a fusion megaproject being built in the south of France by a consortium of 35 nations. It has been in development since the 1980s. Construction began in 2007. The original budget was around $10 billion. The current estimate is north of $22 billion. First plasma was originally scheduled for 2016. Then 2025. Now 2033. Full fusion operations? 2039.
ITER is enormous — 30 meters tall, weighing 23,000 tonnes. It uses conventional low-temperature superconducting magnets that produce about 11-13 Tesla of magnetic field strength. These magnets must be cooled with liquid helium to near absolute zero (4 Kelvin, or -269°C). The engineering is extraordinary. The delays are legendary. Two 11-meter-tall vacuum vessel sections shipped from South Korea didn’t meet millimeter precision requirements. France’s nuclear regulator halted construction. The whole thing has become a cautionary tale about what happens when 35 countries try to agree on how to build anything.
Now here’s what CFS did instead.
In September 2021 — while ITER was wrestling with its latest round of delays — a team of physicists and engineers in a warehouse in Massachusetts tested a prototype magnet made from REBCO tape. It hit 20 Tesla. That’s roughly twice the field strength of ITER’s magnets.
Why does field strength matter? Here’s the analogy.
Imagine you’re trying to hold a water balloon together with rubber bands. The hotter the water (the plasma, in fusion terms), the harder it pushes outward. You need stronger rubber bands to contain it. ITER’s approach is to use a lot of weak rubber bands wrapped around a very large balloon. CFS’s approach is to use much stronger rubber bands wrapped around a much smaller balloon. Same physics. Fraction of the size.
The numbers are dramatic:
| SPARC (CFS) | ITER | |
|---|---|---|
| Magnet strength | ~20 Tesla | ~11-13 Tesla |
| Reactor volume | ~1/40th of ITER | Massive |
| First plasma target | 2026 | 2033 |
| Budget | ~$2 billion (private) | $22+ billion (public) |
| Construction started | 2021 | 2007 |
CFS stayed lean. They hired passionate young physicists straight out of MIT. They thought like a startup, not a government agency. They designed for mass manufacturing from the start — their follow-on power plant design, called ARC, is intended to be built in factories and shipped to sites, not hand-assembled over decades like ITER.
By January 2026, the first of SPARC’s 18 toroidal field magnets was complete and sitting on an assembly jig in Devens, Massachusetts. The machine is being assembled right now.
A Tragic Loss
The fusion community was shaken in December 2025 when Nuno Loureiro — the brilliant Portuguese plasma physicist who had succeeded Whyte as director of MIT’s Plasma Science and Fusion Center — was shot and killed in the foyer of his apartment building in Brookline, Massachusetts. He was 47.
The shooting was later linked to a broader tragedy: the same gunman was responsible for a mass shooting at Brown University three days later. The President of Portugal called Loureiro’s death “an irreplaceable loss for science.”
The work at MIT continues. But the loss is felt deeply. Fusion is a field built by people who dedicate their lives to a problem they may not live to see solved. Losing one of its brightest lights to senseless violence is a wound that doesn’t heal easily.
The Donut vs. The Smoke Ring: Tokamaks and FRCs
Now let’s talk about the other big player in this story: TAE Technologies, and why their approach to fusion is fundamentally different from everything we’ve just described.
CFS builds tokamaks — donut-shaped reactors where plasma spirals around in circles, confined by external magnets. Think of a tokamak as a racetrack for superheated gas, with enormous magnets lining the walls to keep the plasma from touching anything solid (because at 150 million degrees, it would vaporize whatever it touched).
TAE builds something called a Field-Reversed Configuration, or FRC. And it’s a completely different animal.
Imagine blowing a smoke ring. That little donut of spinning air holds itself together — it’s self-contained, self-stabilizing, drifting through space under its own momentum. An FRC is like a smoke ring made of plasma, but instead of floating in air, it sits inside a straight cylindrical tube.
Here’s the key difference: in a tokamak, you provide the magnetic cage. You build enormous magnets around the donut and use them to squeeze the plasma from the outside. The plasma is a prisoner. The magnets are the jail.
In an FRC, the plasma builds its own cage. The electrical currents flowing inside the plasma generate their own magnetic field, and that field is what holds everything together. The external magnets are just there for gentle shaping and stability — like bumper rails on a bowling lane. The plasma is doing most of the work itself.
This has huge practical implications:
- Fewer magnets = less REBCO tape = less supply chain risk
- Cylindrical shape = simpler to build, easier to maintain (you can open it up from the ends)
- Compact = smaller footprint, lower cost
- Self-confining = TAE claims up to 100x more fusion power per unit volume compared to a tokamak at the same field strength
Think of it this way: a tokamak is like building a bank vault around the plasma. An FRC is like teaching the plasma to be its own vault.
Clean Fuel: Why TAE Said No to Radioactive Tritium
Here’s where TAE really zigs while everyone else zags.
Almost every other fusion project on Earth — CFS, ITER, NIF, you name it — uses deuterium-tritium (D-T) fuel. Deuterium is easy to get (it’s in seawater). Tritium is… a problem. It’s radioactive, it’s fantastically expensive, the entire world supply is about 25 kilograms, and when D-T fusion reactions happen, they spray out high-energy neutrons that smash into the reactor walls, making them radioactive over time and requiring heavy shielding.
D-T is the “easy mode” of fusion because it ignites at the lowest temperature — about 150 million degrees. That’s already ten times hotter than the center of the sun, but it’s the most achievable target.
TAE chose hydrogen-boron fuel (proton-boron-11, or p-B11). This is what physicists call an aneutronic reaction — it produces almost no neutrons. Instead of spraying damaging radiation everywhere, the p-B11 reaction mostly produces helium nuclei (alpha particles) that carry electrical charge and can be captured directly as electricity. No steam turbines. No radioactive waste. No tritium.
The catch? Hydrogen-boron requires temperatures of 1 to 3 billion degrees. Not million. Billion. That’s roughly 10-20 times hotter than D-T fusion requires.
This is why TAE needed the FRC. The self-confining plasma geometry, combined with powerful neutral beam injection (think: shooting streams of high-speed atoms into the plasma to heat it up), is how they plan to reach those insane temperatures. In April 2025, TAE achieved a major breakthrough: forming a stable FRC plasma using only neutral beam injection, eliminating an entire layer of startup complexity and cutting projected costs by 50%.
The Trump Media Deal
In December 2025, Trump Media & Technology Group (the company behind Truth Social) announced a $6 billion all-stock merger with TAE Technologies. The deal gives TAE up to $300 million in immediate cash and creates one of the first publicly traded fusion companies.
DJT stock jumped 40% on the news.
The strategic logic, stripped of politics: TAE needed capital and political access. The Department of Energy had been shifting funding away from ITER and toward private American fusion companies. The FY2026 budget cut ITER funding by $45 million while expanding support for private ventures. The merger gives TAE a direct line to the administration and positions it for the kind of government partnership deals — equity stakes, guaranteed purchases, price floors — that have already been structured for companies like MP Materials in the rare earth sector.
TAE’s FRC design also happens to align neatly with the administration’s supply chain strategy. Because FRC requires far fewer superconducting magnets than tokamaks, it’s less vulnerable to the REBCO tape bottleneck and less dependent on Chinese superconductor manufacturing — a significant consideration when you’re simultaneously waging trade wars on multiple fronts.
The plan: begin construction on a 50-megawatt utility-scale fusion power plant in 2026. If everything goes right, commercial fusion power by the early 2030s.
The Full Zoo: Every Way Humans Are Trying to Build a Star
Tokamaks and FRCs aren’t the only games in town. Here’s a quick tour of how everyone else is trying to crack fusion:
Laser Fusion (Inertial Confinement)
How it works: Aim 192 giant lasers at a tiny pellet of fuel the size of a peppercorn. The lasers compress the pellet so fast and so hard that it implodes, crushing the fuel to conditions hotter and denser than the center of the sun. Fusion happens in a flash — literally billionths of a second.
Who’s doing it: The National Ignition Facility (NIF) at Lawrence Livermore. In December 2022, they achieved ignition for the first time. By April 2025, they hit a record 8.6 megajoules of fusion energy from 2 megajoules of laser energy — more than quadrupling their input.
The catch: The lasers themselves consume about 300 megajoules to produce that 2 megajoules of light. The fusion target sees a gain; the electric bill does not. Commercial laser fusion requires dramatically more efficient laser drivers.
Stellarators
How it works: Like a tokamak (donut-shaped), but instead of using a plasma current to help confine itself, a stellarator uses insanely complex, twisted magnetic coils shaped like modern art sculptures. The advantage: inherently steady-state operation with no risk of the sudden plasma disruptions that plague tokamaks.
Who’s doing it: Germany’s Wendelstein 7-X has been setting records. Startups like Proxima Fusion and Type One Energy are designing commercial versions using HTS magnets.
The catch: Those coils are extraordinarily difficult and expensive to manufacture. Every one is a unique shape.
Z-Pinch
How it works: Run a massive electrical current through a column of plasma. The current’s own magnetic field squeezes (“pinches”) the plasma inward, compressing and heating it. No external magnets needed — one of the simplest possible fusion designs.
Who’s doing it: Zap Energy in Seattle. In November 2025, their FuZE-3 device achieved pressures of 1.6 gigapascals — 10 times the pressure at the bottom of the Mariana Trench. They stabilize the otherwise unruly pinch using “sheared flow” — plasma moving at different speeds along the column, like layers of a river flowing past each other.
Magnetized Target Fusion
How it works: A hybrid approach. Confine a plasma target magnetically, then physically crush it with synchronized mechanical pistons driving a liquid metal liner. It’s fusion meets internal combustion engine.
Who’s doing it: General Fusion in Vancouver. Think of a giant mechanical heart, squeezing plasma to fusion conditions with each pulse.
Projectile Fusion
How it works: Fire a hyper-velocity projectile at a fuel target. The kinetic energy of impact triggers fusion. No magnets. No lasers (in the traditional sense). Just very fast things hitting other things very hard.
Who’s doing it: First Light Fusion in Oxford, UK. They’ve crossed $100 million in funding and claim to have achieved fusion conditions using their approach.
Electrostatic Fusion (Fusors)
How it works: Use high-voltage electric fields to accelerate ions toward a central point. Some of them collide and fuse. Simple enough that talented high schoolers have built working fusors in their garages.
The catch: Net energy production is not achievable with this design. But fusors are excellent neutron sources for medical isotopes and materials testing.
The Silver Thread
All of these machines — tokamaks, FRCs, laser systems, stellarators, Z-pinches — share one thing in common. They all need to connect to the electrical grid. And the grid needs silver.
Silver has the highest electrical conductivity of any element. It has the highest thermal conductivity of any metal. It is irreplaceable in high-performance electrical contacts, switching systems, thermal interfaces, and — critically — in the REBCO superconducting tape itself, which uses a silver cap layer to stabilize the ceramic superconductor beneath.
Solar panels need 15-25 grams of silver each. Electric vehicles use 25-50 grams. Data centers consume millions of ounces annually for connectors and thermal management. The grid infrastructure that will deliver fusion power uses silver at every junction, every switch, every protection system.
And here’s the kicker: 72% of silver production comes as a byproduct of mining other metals — copper, lead, zinc. You can’t just “mine more silver” when prices rise. Silver supply is structurally inelastic. The world has run supply deficits for four consecutive years. Silver prices surged 147% in 2025.
In November 2025, the U.S. Geological Survey quietly added silver to the official critical minerals list.
Whether fusion arrives via tokamak, FRC, laser, or something nobody has thought of yet — silver is the connective tissue that makes it all work. It’s the “everything metal” in a world where everything requires electricity.
The Race
So here we are, in February 2026, with more serious fusion projects running simultaneously than at any point in human history:
- CFS assembling SPARC in Massachusetts, targeting first plasma this year
- TAE building toward a 50 MW power plant, freshly capitalized by its Trump Media merger
- ITER grinding forward in France, 18 years into construction
- NIF setting yield records with laser shots
- Zap Energy hitting gigapascal pressures in a device small enough to fit in a shipping container
- Proxima, Type One, Thea designing next-generation stellarators
- General Fusion building a mechanical compression demo plant
The question isn’t whether fusion will work — it already does, in bursts and flashes and record-breaking shots. The question is whether anyone can make it cheap, reliable, and connected to the grid before the world’s energy infrastructure buckles under the weight of AI data centers, electric vehicles, and a climate that’s running out of patience.
The professor gave his students a homework assignment. A few of them started a company. A tape that could carry current without resistance made the impossible merely very difficult. And now, a decade later, the race to bottle a star is the most exciting engineering competition on the planet.
The starting gun has fired. The runners are on the track.
Let’s see who gets there first.
For the full geopolitical and financial analysis — including the REBCO supply chain bottleneck, the silver market repositioning, the Trump administration’s critical mineral strategy, and the Donroe Doctrine — read the complete essay: The Silver Standard