Pulsar Fusion achieves the world's first fusion rocket plasma: what Sunbird proves

2 megawatts of fusion power, aimed at deep space. That is the payload Pulsar Fusion promises the Sunbird drive will deliver, with a specific impulse of 10,000 to 15,000 seconds, or roughly 25 to 40 times what a chemical rocket produces.

In March 2026, the Bletchley-based company lit the first plasma inside the exhaust test rig of that engine. The test ran in the UK, was streamed live to Amazon's MARS Conference in Ojai, California, and announced a simple claim: no fusion rocket exhaust system had done this before.

Nobody has yet flown a fusion engine. The plasma test is a ground rig, and the company's own CEO, Richard Dinan, calls the full product a long way off.

Why spacecraft are stuck between two bad options

Deep-space travel has been a trade-off for six decades. Chemical rockets produce enough thrust to escape Earth's gravity, but their exhaust velocity caps out near 4.5 kilometers per second. That ceiling is why a Mars mission takes seven to ten months and a Pluto probe takes the better part of a decade. New Horizons needed 9.5 years to reach Pluto.

Electric propulsion, the ion and Hall-effect thrusters running on most modern satellites, inverts the problem. Exhaust velocities of 30 kilometers per second and beyond are routine, but thrust is measured in grams. Acceleration is patient to the point of glacial. A Hall thruster can reposition a satellite; it cannot push a crewed ship out of low Earth orbit.

Fusion propulsion is the attempt to combine both. High thrust, because the energy density of the reaction is enormous. High exhaust velocity, because the plasma leaves the nozzle fast. Pulsar's figures for the Sunbird are 10 to 100 Newtons of thrust with exhaust velocity of 98 to 147 kilometers per second.

Those numbers matter, because they change the arithmetic of the solar system.

Mars transit drops from 210-plus days to roughly 150. Pluto, the 9.5-year voyage, becomes a four-year trip. Every day a spacecraft spends in transit is a day it burns fuel and exposes electronics to radiation, so cutting that time has a direct effect on mission cost, not just convenience.

What the first plasma actually proved

The March test was deliberately narrow. Pulsar ran its Mark I exhaust rig with krypton, an inert gas chosen for stable ionization, and demonstrated that electric and magnetic fields could confine plasma inside the Sunbird's exhaust channel and guide charged particles through it. The reaction itself was not a fusion burn. The milestone was the architecture, not the fuel.

That distinction matters for reading the news correctly. A fusion rocket requires four things to come together: a fusion reaction that releases more energy than it consumes, a way to direct that energy into exhaust, materials that survive the radiation, and a power system compact enough to fly. First plasma touches the exhaust channel, the second of those four.

The next phase is where the engineering gets real. It plans to instrument the rig with a thrust balance, E×B probes and retarding potential analyzers to measure actual thrust and exhaust velocity instead of just confirming confinement. The magnetic system is scheduled to move to rare-earth, high-temperature superconducting magnets, and the program ultimately intends to test aneutronic fuel cycles.

The Sunbird architecture: thrust and power from one drive

The Sunbird is designed as a migratory transfer vehicle, not a launch rocket. It is meant to live in orbit, dock with spacecraft that reach low Earth orbit on conventional rockets, and carry them the rest of the way. A ship that would otherwise need 11.3 kilometers per second of delta-v to reach Mars needs only the 9.4 kilometers per second to get to LEO, because the Sunbird covers the trans-Mars injection and Mars orbit insertion.

That cut in delta-v is the entire business case. It estimates the approach reduces total mission delta-v by 20 to 50 percent depending on destination, and slashes launch propellant mass by roughly half. A Mars logistics flight that needs 14.5 kilometers per second of total velocity drops to 9.4. The launch vehicle gets smaller, which gets cheaper, which makes frequent flights plausible.

It sees three commercial mission classes for the tug. Cargo delivery to Mars, 1,000 to 2,000 kilograms in under six months. Outer-planet probes, where high power at the destination enlarges what a science mission can carry. And a lunar orbital supply hub, ferrying supplies to commercial stations. All three share the same logic: the delta-v burden moves from the launch vehicle to a reusable orbital asset.

The power side is the quieter story

Propulsion gets the headlines, but the 2 megawatts of electrical output may be the more durable commercial claim. A spacecraft that arrives at Mars or the outer planets with a compact fusion generator does not depend on weak sunlight. Solar irradiance at Mars is less than half of what it is at Earth; near Jupiter it falls to under 4 percent. A fusion-powered ship carries its own power station, which changes what instruments, relays and life support can do.

For commercial operators, the power output is a product in itself. Data relay, in-orbit manufacturing, fuel electrolysis from water ice, all of these become viable at megawatt scale. The same drive that moves the cargo powers the workshop.

There is a deliberate sequencing here. It sells Hall-effect thrusters today, the LeoBear, Moonranger and Marsranger line, from 500 watts to 10 kilowatts, under ESA and UK Space Agency programs. That is revenue while the fusion program matures. The thrusters are also the station-keeping and maneuvering system that a Sunbird tug would use in orbit, so the commercial business is the flight path to the fusion product.

Where the money went: $14.24 billion into fusion

Fusion funding hit a record in 2025-2026. The Fusion Industry Association's July 2026 report puts cumulative private investment in fusion at $14.24 billion, with $4.5 billion raised in the year to date, nearly double the $2.3 billion of the prior year. The industry's own claim is that this level of capital puts commercial fusion power on a trajectory for the early 2030s.

Almost all of that capital is chasing terrestrial power. Commonwealth Fusion Systems, Helion, TAE and their peers are building reactors to feed the grid, often with hyperscalers as anchor customers. As we wrote in August, data-center operators are now buying fusion power a decade early because compute demand has outrun the buildable supply of dispatchable electricity.

Fusion propulsion sits on the same physics but a different market. The addressable space is smaller, the capital in the sector is a rounding error next to the grid business, and yet the technology roadmap is arguably better defined, because a space engine has a finite power target rather than a net-gain-to-grid gauntlet.

Mars in six months, or Mars in a decade

The gap between the marketing and the schedule is the part a principal needs to weigh. Pulsar's stated mission profile, 150-day Mars transit, assumes the DDFD performs at its design point. The honest baseline is the milestone sequence: first plasma in 2026, in-orbit demonstration of core components in 2027, a production fusion rocket in the early 2030s. Between each of those steps sits an engineering unknown with a history of expanding.

Terrestrial fusion has spent fifty years being a decade away. Space fusion is not exempt from that curse, but it has one structural advantage: a spacecraft engine needs far less energy than a grid reactor. The threshold for useful propulsion is lower than the threshold for net electricity, which is why pulsed and direct-drive concepts can plausibly fly before power plants.

The nearest competition is thin. Helicity Space, backed by Lockheed Martin Ventures and Airbus Ventures, is developing a pulsed magneto-inertial fusion drive with a 2026 plasma-jet demonstration on its roadmap. Princeton Satellite Systems works on a compact fusion concept for propulsion and power. Neither has achieved a live first-plasma demonstration in a flight-class exhaust architecture. It is first to that particular milestone, which is real but early.

Signals to track between now and 2027

For an investor, the question is not whether fusion propulsion works. Physics suggests it can, and capital is finally flowing. The question is whether the 2027 demonstration stays on the calendar, and whether the measured thrust numbers from the instrumented rig support the marketing figures. Those two data points, more than any press release, will separate working technology from a well-funded promise.

The in-orbit economy that fusion propulsion unlocks

The commercial logic of a fusion tug only makes sense if there is a market for fast, heavy transport between orbits. That market is forming faster than most space-infrastructure projections expected. The World Economic Forum and McKinsey put the space economy on a path to triple within a decade, past $1.8 trillion by 2035, with logistics and in-orbit services among the fastest-growing lines.

Look at the demand side rather than the hardware. Constellations need replenishment, stations need resupply, orbital manufacturing needs feedstock, and lunar programs need fuel and equipment moved. Today each of those movements is a dedicated chemical launch, priced by the kilogram, timed by the next launch window. A reusable tug stationed in orbit converts that pattern into a scheduled service, the difference between chartering a freighter and buying a cargo lane.

The delta-v ledger makes the comparison concrete. A mission to Mars assembled in low Earth orbit currently needs roughly 11.3 kilometers per second from Earth's surface, of which 9.4 is just reaching orbit. With a Sunbird-class tug, the launch vehicle only carries the payload to LEO and the tug supplies the remaining trans-Mars injection and Mars orbit insertion. It puts the total mission delta-v reduction at 35 percent for a Mars cargo run and 38 percent for a lunar supply leg, with launch propellant mass down by half.

Those percentages translate into launch prices. Halving the propellant requirement does not halve the ticket, but it does change which missions become commercially viable. A 1,000 to 2,000 kilogram cargo run to Mars in under six months at a docking-service price is a different product from a one-off flagship science launch at $50,000 per kilogram. The first makes logistics a recurring business; the second stays an occasional event.

There is an equity lesson in the sequencing. The companies with durable claims in this story are not only the fusion engine builders. The Hall-effect thruster makers, the in-orbit servicing operators, the refueling startups and the orbital logistics platforms all capture value as the transport layer thickens, regardless of which fusion concept wins. Its own structure reflects that, selling thrusters commercially today while the fusion program burns through its milestones.

That is the most investor-relevant part of the whole story. Fusion propulsion is a frontier bet with a long fuse, but the infrastructure it presupposes, reusable tugs, orbital depots, high-power spacecraft, is being built now by companies that do not need fusion to succeed to make money. The transport layer is investable before the engine is proven.

What stays open is the timeline. The 2027 in-orbit demonstration is the hinge. If Sunbird's core components operate in vacuum as designed, the early-2030s production target becomes plausible and the tug economics start to close. If that demonstration slips, the commercial market will have two or three more years to build orbital logistics on electric propulsion alone, and fusion propulsion will have to win later at higher cost.