On June 16, 1992, a 30-meter, 185-ton white catamaran named Yamato 1 pulled out of Kobe harbor. It made no wake behind a propeller — because it had no propeller. It had no driveshaft, no gearbox, no rudder linkage to a screw. Seawater flowed through two thruster tubes, was hit with a magnetic field and an electric current, and was flung out the back by the Lorentz force. The ship moved. It was the first — and still only — full-scale vessel ever propelled by magnetohydrodynamics.

The idea traces to Yoshiro Saji of Kobe University, who published on superconducting MHD ship drive in 1976. The Japan Foundation for Shipbuilding Advancement bankrolled the demonstrator in 1985. Mitsubishi Heavy Industries built the hull. Kobe Steel built the six-pole dipole magnets: NbTi superconductors cooled with liquid helium to 4.2 K, producing a 4-tesla field across a bore full of seawater. Electrodes injected current perpendicular to the field. F = J × B. Water accelerated aft. Ship went forward.

It reached 8 knots. And that killed it.

Conventional diesels shove destroyers around at 30+ knots. MHD's problem was brutal thermodynamics: thrust scales as B², and 4 T just wasn't enough to be competitive. Overall efficiency limped in around 2%. The seawater electrolyzed at the electrodes, producing chlorine gas. The cryogenics were a nightmare — liquid helium at $10/liter, complex vacuum plumbing, magnets that quenched if you looked at them wrong. In 2000, the Ship & Ocean Foundation quietly wound the program down. Yamato 1 was moved to the Kobe Maritime Museum, where it sits today as a curiosity.

Here's why 2026 changes the math completely.

• REBCO high-temperature superconductors. Commonwealth Fusion's SPARC tokamak demonstrated a 20-tesla REBCO magnet in 2021. That's 5× Yamato's field. Since thrust scales as B², that's a 25× improvement in force density from the magnets alone. And REBCO operates at 20 K — you can cool it with a closed-cycle cryocooler running off ship power. No liquid helium logistics.
• Electrode chemistry is solved. Modern platinum-group and diamond-coated electrodes suppress chlorine evolution. The Chinese Academy of Sciences has published on catalyst coatings that push MHD electrical efficiency past 90% at the electrode interface.
• China is already back in the water. In 2017, the Naval University of Engineering in Wuhan tested a small submersible MHD demonstrator. By 2022 they had a 1-ton test rig running at 3.5 T. The strategic logic is obvious: MHD drives are acoustically silent. No cavitation, no gearbox whine, no shaft-line vibration. A nuclear submarine's biggest passive-sonar signature is its screw.
• Efficiency now favors specific missions. Even at 40–50% overall efficiency (achievable at 20 T), MHD isn't competing with container ships. It's competing with silent submarines, autonomous underwater vehicles that need to run for months without seals wearing out, and shallow-water craft where a spinning propeller kills manatees or eats mud.

The Yamato 1 was cancelled for being slow. That was true in 1992 with NbTi at 4 T. In 2026, with REBCO at 20 T and stealth as a design driver instead of speed, it looks less like a failed prototype and more like a working demonstrator of a technology that just needed the magnets to catch up.