Pacific Fusion is building a pulsed-power fusion system that operates more like an internal combustion engine than a conventional reactor. Instead of maintaining a continuously burning plasma inside a magnetic bottle, banks of capacitors store electrical energy and release it in a burst lasting roughly 100 nanoseconds. The resulting magnetic pressure crushes a small cylindrical target until its deuterium-tritium fuel fuses. A future power plant would repeat this cycle approximately once per second, capture the heat in a surrounding blanket, and generate electricity through a conventional thermal power cycle.

The core hardware is the impedance-matched Marx generator, or IMG, a modular pulser assembled from repeating units called bricks, each containing two capacitors and a switch. Bricks are combined into stages and shipping-container-sized modules. The architecture converts stored capacitor energy into an electrical pulse in a single step, removing the intermediate pulse-compression stages used in traditional pulsed-power machines. Pacific Fusion claims roughly 90% conversion efficiency from stored energy to the initial electrical wave.

In June 2026, Pacific Fusion demonstrated a one-third-scale prototype module that produced approximately 440 gigawatts of peak power and 1.1 megavolts in an 80-nanosecond pulse, with more than 1,000 qualification shots completed. Sirius, a related prototype built at Lawrence Livermore National Laboratory, exceeded 3,000 full-power shots by July 2026, providing initial data on the IMG architecture's ability to fire repeatedly with controlled performance.

The Demonstration System under construction in Albuquerque will combine 156 modules to deliver more than 60 mega-amps to a central target, with the goal of producing over 100 megajoules of fusion energy and achieving net facility gain by 2030. The system is also intended for national-security testing, generating intense neutron, X-ray, gamma-ray, and electromagnetic environments for stockpile stewardship, survivability testing, and high-energy-density physics.

The fuel target is a small cylindrical assembly of plastic and aluminum containing magnetized deuterium-tritium fuel. Pacific Fusion tested self-magnetizing target concepts on Sandia's Z machine in early 2026, removing the need for an external magnetic coil that would be destroyed in every shot. Because each target is consumed during the fusion event, commercial operation will require low-cost, high-volume target manufacturing.

Pacific Fusion is a vertically integrated deep-tech infrastructure company serving energy and national-security markets. It designs and manufactures pulser modules, develops fusion targets and simulation tools, and is building demonstration and manufacturing facilities. Its pulser architecture uses commodity materials such as steel, aluminum, plastics, and oil, with the goal of enabling factory production rather than bespoke on-site construction.

The near-term business-to-business strategy targets government agencies, national laboratories, defense contractors, and research institutions. Starting in 2029, the Demonstration System is intended to host paid experimental campaigns, government-funded research programs, and facility-access agreements. Pacific Fusion signed a memorandum of understanding with the National Nuclear Security Administration in August 2026 covering potential collaboration in high-yield fusion and national-security applications, though specific projects and funding require separate agreements.

The long-term commercial model centers on fusion power plants. Potential structures include selling or licensing modular plant systems to utilities, building and operating plants under long-term power purchase agreements, or providing dedicated behind-the-meter generation for data centers, defense installations, and industrial campuses. Pacific Fusion has publicly targeted electricity costs below $0.05 per kilowatt-hour but has not published a validated plant-cost model.

Milestone-tranched financing determines the company's cost profile, with capital released against technical achievements to align spending with technical de-risking and provide multi-year funding for hardware construction and hiring. The roughly $1 billion Albuquerque campus investment, including the Demonstration System estimated at approximately $500 million, is the dominant near-term cost. Operating expenses include specialized scientific and engineering labor, high-voltage testing, large-scale simulation, target fabrication, and national-laboratory partnerships. The company grew from 87 employees in early 2025 to more than 200 by June 2026.

The business model depends on a feedback loop between government, scientific, and commercial energy applications. Government and scientific users could fund shots and component improvements, generating reliability and plasma-performance data that inform simulation and target design and reduce power-plant technical risk. Progress toward fusion power could then increase the facility's value to government users, partially bridging the period between venture-funded R&D and commercial electricity sales.

Pacific Fusion competes across a segmented fusion market in which scientific validation, plant engineering, commercial credibility, and access to capital matter alongside the timing of the first fusion demonstration.

Zap Energy is the closest competitor in electrically pulsed, magnetically compressed fusion without superconducting magnets or lasers. Its sheared-flow-stabilized Z-pinch is simpler and more compact, while its Century platform had demonstrated repetitive operation at 0.2 Hz by September 2025. If Zap reaches adequate plasma gain, its smaller unit size and operating experience could translate into a faster, less capital-intensive path to a power plant.

First Light Fusion uses an electromagnetic launcher to accelerate a projectile into a specially designed target, competing for the same inertial-fusion scientists, target-fabrication capabilities, and government programs. Its asset-light licensing model may require less capital than Pacific Fusion's integrated campus strategy. Laser-driven companies including Focused Energy, Xcimer Energy, and Marvel Fusion are building on NIF's demonstrated ignition but must reduce laser-driver costs while increasing efficiency and firing frequency.

Commonwealth Fusion Systems competes with Pacific Fusion for institutional backing and utility customers. Its SPARC tokamak targets net fusion energy in 2027, and Google has agreed to buy half the output of its planned 400 MW ARC plant. CFS has also vertically integrated high-temperature superconducting magnet manufacturing, giving it control over a technical and manufacturing capability available to few competitors.

Helion Energy is the primary time-to-market competitor. It is building Orion in Washington state for Microsoft under a 50 MW power purchase agreement and has a separate collaboration with Nucor for industrial power. Helion's direct electricity conversion approach could eliminate blankets, heat exchangers, and turbines. If Orion delivers meaningful net electricity near Helion's timetable, it could attract customer and investor attention before Pacific Fusion's 2030 facility-gain milestone.

General Fusion is pursuing magnetized-target fusion using mechanical compression and a liquid-metal wall, with its LM26 machine operating by early 2026. TAE Technologies targets hydrogen-boron fuel, which could reduce neutron damage and tritium dependence if it can achieve the materially more demanding plasma conditions.

Pacific Fusion also competes with technologies that do not require an unresolved plasma-physics breakthrough. Advanced fission small modular reactors target the same demand for reliable, carbon-free power from data centers and defense installations, while natural gas with carbon management remains dispatchable and fast to construct. Enhanced geothermal and renewables-plus-storage continue to improve in cost and availability, requiring fusion plants to compete on total delivered cost rather than fuel economics alone.

Pacific Fusion's TAM expansion strategy extends beyond grid electricity to national security, research infrastructure, and, eventually, global power markets through its dual-use pulser platform.

The nearest-term expansion opportunity is selling access to extreme experimental environments before a commercial power plant exists. Pacific Fusion's high-yield system can generate neutron, X-ray, gamma-ray, and electromagnetic conditions relevant to nuclear-stockpile stewardship, satellite and electronics survivability testing, radiation-effects qualification, and high-energy-density physics. The August 2026 NNSA memorandum of understanding provides a framework for collaboration in these areas.

No other U.S. facility currently under construction combines Pacific Fusion's planned output, fluence, and private-sector operating model. Existing government facilities such as NIF and Sandia's Z machine face recapitalization requirements and constrained access, leaving demand for a privately operated, high-yield pulsed-power platform that could host experimental campaigns for government agencies, defense contractors, and research institutions.

AI-driven data-center power demand is projected to reach 123 GW by 2035, while U.S. electricity consumption is expected to grow close to 2% annually through 2030, with data centers accounting for roughly half of incremental demand. These concentrated loads place a premium on firm, high-capacity generation and could create anchor customers willing to enter long-dated power-purchase or development agreements.

Modular fusion plants could serve data-center campuses, advanced manufacturing clusters, military installations, and remote industrial sites that require firm power without fossil-fuel exposure. Behind-the-meter or campus-scale deployments, where customers value reliability, land efficiency, and energy security over the lowest wholesale price, could reduce first-of-a-kind commercial risk.

Pacific Fusion's architecture, based on 156 nominally identical pulser modules, creates a potential industrial-equipment business independent of owning every power plant. The company could sell complete pulsers, replacement modules, switching systems, controls, and maintenance services to government laboratories, defense contractors, and other fusion developers.

The Fusion Factory in Los Lunas and the San Leandro Build Center provide early manufacturing capacity and proprietary process knowledge. Over time, standardized modules could be used for upgrades, fleet maintenance, and adjacent applications in radiation testing, materials science, and advanced manufacturing, extending the fusion driver into a broader industrial product line.

Repetition-rate gap: Moving from occasional research shots to approximately one shot per second requires target insertion, vacuum recovery, debris management, chamber clearing, and component survival at a cadence four to six orders of magnitude higher than at current major research facilities, none of which Pacific Fusion has demonstrated as an integrated system at commercial scale.

Consumables economics: Because every fusion shot destroys a target and may destroy nearby transmission-line sections, a one-hertz commercial plant would require more than 31 million precision-manufactured targets annually before rejects or downtime, making individually inexpensive consumables a material operating cost at scale.

Customer-development lag: Pacific Fusion has no announced power purchase agreement, named commercial electricity customer, or binding government procurement contract, and its mid-2030s commercial power timeline may follow commitments by hyperscalers and utilities to competing generation platforms from companies like Commonwealth Fusion Systems and Helion Energy that are securing customers today.