Nuclear Fusion
The energy of the stars on Earth: the holy grail of the energy transition
Overview
Beyond ITER: private startups and public projects are closing in on commercial Q > 1. Tokamaks, inertial confinement, stellarators and alternative approaches.
Impact: Potential low-carbon energy source, still dependent on scientific and industrial breakthroughs
Involved actors
Projects
Development
Commonwealth Fusion Systems — SPARC
Commonwealth Fusion Systems
Private-sector leader. $4B raised in total (July 2026). Record HTS magnets. ARC = commercial reactor.
Compact tokamak with HTS (high-temperature superconducting) magnets. SPARC is targeting net energy gain (Q > 1) in 2027, the threshold at which the machine produces more fusion energy than it consumes.
- Compact tokamak
- HTS magnets
- Plasma
- Commercial power
Roadmap: SPARC first plasma
2027
Helion Energy — Polaris
Helion Energy
The only startup with a pre-signed customer: Microsoft, for commissioning in 2028 then at least 50 MW. Non-tokamak approach.
FRC (field-reversed configuration) approach. In February 2026, Polaris became the first privately developed fusion machine to demonstrate measurable deuterium-tritium fusion. The Orion plant is under construction in Malaga, Washington.
- FRC
- Direct energy conversion
- Deuterium-Helium-3
- Commercial PPA
Roadmap: Orion: commissioning of the first fusion power plant
2028
TAE Technologies — Norm then Da Vinci
TAE Technologies
Aneutronic approach (proton-boron 11). One machine generation skipped. Google among the investors.
Beam-driven FRC reactor. Norm, the first machine to produce FRC plasmas through neutral beam injection alone, exceeded its objectives: in November 2025, TAE removed its planned sixth machine, Copernicus, from its roadmap to go straight to Da Vinci, its first power plant.
- FRC
- Proton-Boron 11
- Aneutronic
- AI-guided plasma
Roadmap: Da Vinci: net energy gain and first power onto the grid
2030+
Zap Energy — Sheared-flow Z-pinch
Zap Energy
A radically different approach. FuZE-Q prototype under construction.
Shear-stabilized Z-pinch approach. No magnets, no lasers — the simplest and cheapest design.
- Z-pinch
- No magnets
- Compact
- Low-cost fusion
Roadmap: FuZE-Q: net gain demonstration
2028