Realta Fusion Unveils Hammir-DT, a Tandem Magnetic Mirror Power Plant Design Targeting 500 MW and Sub-$50/MWh Electricity
The Madison startup detailed a 500 MW deuterium-tritium magnetic mirror plant using REBCO magnets from Commonwealth Fusion Systems, claiming a $50/MWh path.
Overview
Realta Fusion, the Madison, Wisconsin startup pursuing magnetic mirror fusion, has detailed Hammir-DT, a preconceptual design for a deuterium-tritium power plant it describes as the first modern axisymmetric tandem mirror power plant design. According to The Fusion Report, the design targets 500 MW of fusion power, 538 MW thermal, and approximately 200 MWe delivered through direct energy conversion. The company has put a cost envelope on the concept that few fusion developers have been willing to state publicly: a capital expenditure of $1 billion to $2 billion per plant and a levelized cost of electricity below $50 per megawatt-hour.
The announcement extends a year in which Realta has moved from validating the physics of its approach to sketching a commercial machine. It also deepens the company’s reliance on Commonwealth Fusion Systems, whose high-temperature superconducting magnets sit at the center of the Hammir-DT design.
What We Know
The mirror architecture is what sets Realta apart from the tokamak and stellarator programs that dominate private fusion. Rather than confining plasma in a closed ring, a tandem mirror traps it along a straight central cell capped at each end by high-field magnetic “plugs.” Per The Fusion Report, Hammir-DT uses four high-field magnets per plant, two at each end, while the central cell magnets operate at 1 to 3 Tesla. The plant is designed to run continuously rather than in pulses, which the company frames as an advantage over pulsed-power fusion concepts.
The high-field magnets come from a partnership Realta and Commonwealth Fusion Systems announced earlier this year. According to the companies’ joint statement on April 2, 2026, CFS will develop high-temperature superconducting (HTS) magnets for Realta’s demonstration prototypes and commercial fusion power plants. “By working with the world’s leading HTS magnet manufacturer, we are significantly de-risking one of the most critical of these technologies,” said Kieran Furlong, Realta’s chief executive. Bob Mumgaard, CFS chief executive and co-founder, said, “As the world leader in HTS magnet technology development and manufacturing, we are pleased to share our expertise with Realta.” CFS is based in Devens, Massachusetts.
Hammir-DT relies on REBCO high-field magnets from Commonwealth Fusion Systems and on direct energy conversion to turn the energy of charged fusion products into electricity, as reported by The Fusion Report. The design also counts on tritium breeding: the extra 38 MW separating the 500 MW of fusion power from the 538 MW thermal figure comes from the exothermic energy released when breeding tritium from lithium-6. The company’s stated design goals include a scientific gain QSCI of 5.3 and an engineering gain QENG of 1.3.
Those figures build on modeling Realta published as it spun out of academic research. According to Power Online, the company’s earlier work projected an energy gain of Q greater than 5 with a 50-meter center cell, and potentially Q greater than 10 or higher with a longer center cell. “Our model proves there’s a there there when it comes to CoSMo fusion energy systems based on the tandem mirror,” said Dr. Sam Frank, the company’s computational physics team leader.
The company’s roots trace to a University of Wisconsin-Madison machine. Per the Realta-CFS statement, ARPA-E funded the university to build the Wisconsin HTS Axisymmetric Mirror (WHAM) experiment in 2020, Realta spun out of WHAM in 2022, and WHAM achieved first plasma confinement at 17 Tesla in 2024. Realta is backed by Khosla Ventures and Future Ventures and was selected for the Department of Energy’s fusion milestone program. According to Power Online, it was among eight organizations chosen for the Department of Energy’s Milestone-Based Fusion Development Program in 2023 and earlier closed a $36 million Series A round led by Future Ventures with continued backing from Khosla Ventures. The Hammir-DT design positions the company to stack center-cell modules for outputs ranging from 100 MW to 500 MW per unit, according to The Fusion Report.
What We Don’t Know
Hammir-DT remains a preconceptual design, not a funded construction project. Realta has raised $54 million to date, as reported by The Fusion Report — a fraction of the $1 billion to $2 billion per-plant capital cost the design itself implies, meaning the company will need a far larger raise to build a first commercial unit. The published gain figures are modeled targets rather than demonstrated results, and the leap from WHAM’s experimental plasma to a continuously operating 500 MW reactor spans technical steps the company has not yet completed. No construction timeline or site for a Hammir-class plant has been disclosed.
Analysis
Realta’s willingness to attach hard cost numbers — a per-plant CapEx band and a sub-$50/MWh levelized cost target — is unusual in a field where most developers emphasize physics milestones over economics. If the figures hold, they would place mirror fusion within range of conventional generation costs. The bet also tightens the link between two complementary companies: Commonwealth Fusion, which this month published five peer-reviewed papers validating the physics of its 400 MW ARC tokamak, supplies the HTS magnets that make Realta’s high-field mirror plugs feasible. The same magnet technology now underpins two distinct confinement strategies, a sign of how central HTS magnet manufacturing has become to the private fusion race. Whether Realta can convert a preconceptual paper into financed steel remains the open question.