Pacific Fusion
Pre-revenue deep-tech: venture-funded R&D toward a demonstration fusion system, with an early open-access users program (external researchers, commercial applications, national-security experiments) as a path to first monetization before commercial power plants
Pacific Fusion has never disclosed a round valuation; the valuationB figures are cumulative committed-capital proxies (a strict floor on post-money, since post-money >= capital raised), not priced equity marks. A priced Series B would provide the first true external valuation point.
Earnings, margins, COGS & capex
Private and pre-revenue; no audited financials are public. The capital story substitutes for the income statement: a ~$900M Series A (Oct 2024, the largest Series A in fusion and among the largest venture Series A rounds on record) structured as milestone-based tranches, with the round now exceeding $1B after the June 2026 pulser-module milestone. Spending is concentrated in pulsed-power module engineering, target physics (experiments at Sandia's Z facility under a CRADA), and the ~$1B New Mexico demonstration campus plus the Los Lunas module factory. First plausible revenue is the Pacific Fusion Users Program (commissioning from 2028; first external user campaigns anticipated 2029); power sales are well past 2030 - the Demonstration System has no turbine and will not generate electricity.
Revenue trend
Margins
n/a
n/a
n/a
COGS structure
Not applicable yet. Future cost structure centers on mass-manufactured pulsed-power 'bricks' (each two capacitors plus one switch) assembled into impedance-matched Marx generator modules, deuterium-tritium targets, and facility operations - the company's core cost thesis is that pulser modules are factory-producible from commodity electronics, unlike bespoke laser or superconducting-magnet systems.
Capex
Dominant cash use: ~$1B Albuquerque (Mesa del Sol) research and manufacturing campus with 200+ long-term jobs and hundreds of construction jobs (announced Sep 26, 2025); Los Lunas, NM build center (opened Dec 12, 2025) manufacturing the 156 pulser modules, targeted for production by roughly end of 2027; Demonstration System construction expected to begin summer 2026.
Latest earnings
Not applicable; the tracked equivalent is technical milestones: in June 2026 the scaled pulser prototype (9 stages, 90 bricks, about one-third of a full module) delivered 440 GW peak power and ~1.1 MV in an 80-ns pulse - performance the company reports as meeting all requirements to scale to the full system - and unlocked a further Series A tranche (a beat, in milestone terms)
Company-stated roadmap: begin Demonstration System construction summer 2026, manufacture 156 full-size modules (32 stages of 10 bricks each) at Los Lunas by roughly end of 2027, commission from 2028, first external user campaigns 2029, and achieve net facility gain (fusion output exceeding total stored energy input; no electricity generation - the demo has no turbine) by 2030
- Series A committed (Oct 2024)
- ~$900M, milestone-tranched; round now exceeds $1B after the Jun 2026 unlock
- Prototype performance (Jun 2026)
- 440 GW peak power, ~1.1 MV, 80 ns pulse (one-third-scale module)
- Pulser modules required
- 156 (prototype validated; full-size modules to be built at Los Lunas by ~end 2027)
- NM campus investment
- ~$1B (Mesa del Sol, Albuquerque)
- Net facility gain target
- 2030
Growth drivers
- Milestone-tranched capital — each technical proof (Feb 2026 simplified-target results from Sandia Z-facility experiments; Jun 2026 scaled pulser prototype at 440 GW / ~1.1 MV / 80 ns) advances the plan, with the Jun 2026 result unlocking a further Series A tranche
- Demonstration System schedule — construction from summer 2026, commissioning from 2028, net facility gain targeted by 2030
- Pacific Fusion Users Program (PFUP, launched Mar 2026) — external access for national labs, academia, and industry - fusion energy science, commercial applications, fundamental science, national security; first external campaigns anticipated 2029
- AI/data-center electricity demand raising the strategic value of firm clean power and hyperscaler willingness to sign early fusion offtakes (precedents: Microsoft-Helion, Google-CFS)
- Supportive US policy — DOE milestone-based fusion programs and state/local incentives (Albuquerque legislation cleared the campus path in 2025)
Bull & bear
Pacific Fusion is the best-capitalized pure pulsed-power fusion play, attacking the approach with the cheapest theoretical path to net gain - Sandia's Z-machine physics lineage with a modular, factory-producible driver - and it is hitting its published milestones on schedule with a capital structure that rewards exactly that.
- Milestone-tranched Series A now exceeding $1B means capital efficiency is enforced, not promised - and the Jun 2026 prototype milestone unlocked its tranche on schedule
- Pulser-driven ICF requires no lasers and no high-temperature superconducting magnets: the driver is commodity capacitors and switches in mass-producible modules, giving a credible path to low $/W plant capex
- Sandia's MagLIF lineage plus NIF's demonstrated ignition de-risk the underlying physics more than most other private inertial approaches, and the Feb 2026 simplified-target result removed the need for expensive single-use external coils
- The Users Program creates pre-power revenue optionality (national-security HED science, commercial applications, fundamental science) that laser and tokamak rivals mostly lack
- Board and investor network (Schmidt, Collison, Taneja, Griffin, Doerr) is a structural advantage for follow-on capital and future hyperscaler offtakes
- A 2030 net-facility-gain result would leapfrog the company into the top tier of fusion valuations at a fraction of CFS's cumulative spend
This is a pre-revenue physics bet with an undisclosed valuation, no offtake customer, and a 2030 net-gain target in an industry where schedules have slipped for seventy years; even flawless execution leaves commercialization and fuel-cycle problems unsolved into the mid-2030s.
- Net facility gain is undemonstrated: one one-third-scale module at 440 GW is not 156 full-size modules synchronized onto a DT target at ignition-relevant conditions - the system-integration risk is the whole company
- No disclosed valuation, revenue, or burn rate: outside investors cannot price the equity; the only marks are tranche unlocks controlled by insiders
- Pulsed ICF power plants need high repetition rates and cheap mass-produced targets; the Demonstration System is single-shot-class science with no turbine, leaving the hardest commercial problems (rep-rate, target factory, tritium breeding) entirely ahead
- Competitors have customers: Helion has a Microsoft PPA (50 MW, delivery targeted 2028) with its Orion plant under construction, and CFS has a 200 MW Google offtake with an announced ARC site in Virginia - Pacific Fusion has expressions of interest for a users program
- A milestone miss in demo construction (starting summer 2026) or module manufacturing ramp could freeze the tranche structure precisely when capital markets for long-duration deep tech are selective
- Fission SMRs (Oklo, NuScale, GE Vernova-Hitachi) and solar+storage are commercially deployable this decade and may absorb the data-center firm-power demand fusion is counting on
What it is worth
Last-round anchoring + milestone-adjusted qualitative scenarios. No public valuation exists: the ~$900M Series A (Oct 2024) valuation was not disclosed, and third-party estimates are unverified. The honest anchor is capital committed (>$1B, milestone-tranched) and comparables: Helion was valued at $15.5B post-money in its Jun 2026 Series G ($465M, led by Thrive Capital) with a Microsoft PPA and a plant under construction; CFS has raised close to $3B (its Series B2 valuation was not officially disclosed). Both comps carry either an offtake customer or an operating integrated prototype, which Pacific Fusion lacks.
A module-manufacturing or demo-construction milestone miss freezes tranches and forces a down-round or strategic sale; in the worst case the General Fusion pattern (capital strain before integrated proof) marks the equity toward the asset value of the NM facilities and IP.
Milestones continue to unlock through demo commissioning (2028) but net gain slips modestly past 2030: valuation grows with de-risking into the billions, with a priced Series B providing the first external mark.
Net facility gain achieved by 2030 on schedule with a validated low-cost module manufacturing curve: re-rates toward the Helion/CFS tier (multi-billion to low-double-digit billions), implying several-fold appreciation from any plausible Series A mark.
Any mark is a probability-weighted bet on 2030 net facility gain. The milestone structure means dilution risk is front-loaded into technical outcomes rather than market timing. Not financial advice; private shares are illiquid and any secondary quotes are indicative only.
SWOT
Strengths
- Largest Series A in fusion (~$900M committed Oct 2024, now >$1B) with milestone discipline that aligns spend to de-risking
- Elite backer and board bench — General Catalyst (Hemant Taneja), Breakthrough Energy Ventures, Ken Griffin, Patrick Collison, John Doerr, Eric Schmidt, Mustafa Suleyman, Andrew Forrest, Lightspeed, Lowercarbon Capital; Collison, Taneja, and Schmidt on the board
- Pulsed-power approach builds on Sandia's Z-machine/MagLIF physics lineage; CTO Keith LeChien co-invented the impedance-matched Marx generator the architecture depends on
- Manufacturability thesis — modular, shipping-container-scale pulser modules from commodity electronics rather than bespoke lasers or HTS magnets
- Demonstrated execution cadence — two public technical results in 2026 (simplified aluminum-and-plastic target validated in Sandia Z-facility experiments, Feb; 440 GW one-third-scale prototype completing the second milestone set, Jun)
Weaknesses
- Pre-revenue with no disclosed valuation, revenue, or burn — investors have no financial anchor, only technical milestones
- Net facility gain not yet demonstrated — the core scale-up (a one-third-scale module to 156 synchronized full-size modules driving DT targets) remains unproven
- Founded 2023 — years behind Commonwealth Fusion Systems and Helion in facility build-out and customer contracts
- No announced power offtake agreements, unlike Helion (Microsoft) and CFS (Google)
- Tritium fuel-cycle handling, target mass-production, and repetition-rate operation are unsolved engineering challenges for any inertial approach
Opportunities
- AI data-center load growth creating hyperscaler demand for firm clean power and early-offtake appetite
- PFUP users program could generate first revenue and institutional lock-in (labs, national security, commercial applications) from 2029, before any power sales
- NIF's 2022+ ignition results validated inertial fusion physics broadly, improving the category's credibility
- US policy tailwinds — DOE milestone programs, New Mexico state and Albuquerque city incentives, and national-security interest in high-energy-density facilities
- If the module cost curve holds, pulsed power could undercut tokamak and laser capex per plant
Threats
- Any missed technical milestone can stall the tranche structure and force a re-raise in a harder market
- Better-funded competitors (CFS with close to $3B raised — Helion at a $15.5B valuation with a signed Microsoft PPA and a plant under construction) could win the talent, supply-chain, and customer race
- Fusion timelines have historically slipped — a 2030 net-gain miss would compress credibility and follow-on funding
- Cheap alternatives (fission SMRs, solar+storage, geothermal) may capture the firm-clean-power market before fusion is commercial
- Regulatory framework for fusion (NRC byproduct-materials approach) is favorable today but could tighten
Moats, dependencies & bottlenecks
Moats
CTO Keith LeChien co-inventor, ex-NNSA/Sandia-lineage talent) Deep bench in a narrow discipline where the global talent pool is small and mostly at national labs
Committed capital through the demo build is rare; but it is contingent on continued technical delivery
Z-facility experiments) and MagLIF physics lineage Working access to the institution with decades of Z-machine data; not exclusive
If module costs fall on a manufacturing learning curve at Los Lunas, this becomes the defining cost moat vs lasers/magnets; unproven today
weak-to-moderate Site control, incentives, and workforce pipeline (UNM partnerships) near Sandia/Los Alamos talent
Dependencies
The entire plan through 2030 assumes milestones keep releasing committed capital; a miss stalls the build
Sandia National Laboratories (target-physics experiments under CRADA at the Z facility; MagLIF heritage) The Feb 2026 simplified-target validation ran on Sandia Z-facility experiments; federal lab access depends on agreements and politics
156 modules (each 32 stages of 10 two-capacitor bricks) require volume supply of high-voltage capacitors and switches; a concentrated supplier base could gate the schedule
DT operations at net-gain scale require tritium inventory and regulatory clearances
regulatory/political low-to-medium City legislation cleared the campus path (2025); construction and operation permits still sequence the 2028 commissioning
Competes with national labs and other fusion startups for a very small global cohort
Advantages
- Cheapest-driver thesis — commodity pulsed-power electronics vs lasers (Xcimer, NIF-style) or HTS magnets (CFS) - potentially far lower plant capex
- Physics head start from decades of public Z-machine/MagLIF data rather than a novel confinement concept
- Capital committed through the demonstration phase, insulating it from venture-market cycles if milestones hit
- Dual-use facility economics — HED science, national security, and commercial applications monetizable via the Users Program before net gain
- Founding CEO Eric Lander's scientific stature (Human Genome Project; former White House OSTP director) and the investor network accelerate policy access and follow-on capital
Weaknesses
- No demonstrated net gain, no revenue, no disclosed valuation - pure milestone-stage risk
- No power offtake customer signed, trailing Helion and CFS on commercial validation
- Later-stage rivals have operating or under-construction integrated devices (Helion's Polaris operating and Orion plant under construction; CFS assembling SPARC); Pacific Fusion's demo breaks ground summer 2026
- Single-site concentration: the entire technical program funnels through one Albuquerque campus
- Inertial approaches must eventually solve repetition-rate operation, which the Demonstration System does not attempt
Bottlenecks
- Demonstration System construction start (summer 2026) and the 2028 commissioning schedule
- Manufacturing ramp of 156 full-size pulser modules at Los Lunas (targeted by roughly end of 2027)
- System-level synchronization of 156 modules onto a target at ignition-relevant current (the one-third-scale 440 GW result does not prove this)
- Mass-producible, low-cost DT target manufacturing for any future rep-rated commercial system
- Tritium fuel cycle: supply, handling, and eventually breeding - unsolved industry-wide
Top signals & trends
Top signals
First visible slip risk on the 2028-commissioning / 2030-net-gain path
The manufacturing learning curve is the cost thesis; slippage gates commissioning
A priced Series B would give the first external valuation mark
First revenue and third-party validation of the facility
Category validation but competitive pressure on future offtakes
Federal support shifts economics for the whole sector
Trends
Creates the future customer base and early-offtake precedent (Microsoft-Helion 50 MW, Google-CFS 200 MW)
NIF's 2022+ ignition results proved the physics; pulser-driven ICF rides that validation
The industry has raised roughly $10B privately; capital now demands staged proof, which suits Pacific Fusion's structure
Oklo, NuScale, and GE Vernova-Hitachi deployments this decade could crowd the market fusion targets for the 2030s
Context for urgency and US policy support for domestic fusion; a sector observation, not an investment call
Materially shortens the path from demo to commercial siting
Ecosystem & competitor graph
Suppliers feed the company; customers pull from it. Line thickness shows the strength of each tie (supply-chain dependency, customer earnings contribution). Hover to isolate a tie.
Representative of the high-voltage electrical infrastructure and power-management component class used in pulsed-power builds (not a confirmed Pacific Fusion supplier)
Representative of the capacitor/passives class the modular pulser thesis depends on (not a confirmed supplier)
Representative film/power capacitor supplier class for Marx-generator stages (not a confirmed supplier)
Private; the dominant US inertial-fusion target fabricator (NIF targets) - the reference supplier class for DT target manufacturing
Representative industrial-gas supplier class including deuterium supply
LLNL, LANL) PFUP demand class for high-energy-density science and national-security experiments; Sandia is operated by NTESS, a Honeywell (HON) subsidiary
Prospective future offtake class - established the hyperscaler fusion-PPA precedent with Helion; not a Pacific Fusion customer today
Prospective future offtake class - signed the 200 MW CFS offtake precedent; not a Pacific Fusion customer today
Representative buyer class if PFUP commercial applications extend to isotope production; no relationship announced
Private; best-funded fusion company (close to $3B raised, ~one-third of all private fusion capital), SPARC tokamak in assembly and a 200 MW Google offtake for its planned ARC plant in Virginia; the sector benchmark
Private; pulsed field-reversed-configuration approach, $15.5B post-money valuation (Jun 2026 Series G), Microsoft PPA, Polaris prototype operating, Orion plant under construction in Washington; closest analog in pulsed fusion with a commercial head start
Private; sheared-flow-stabilized Z-pinch - the other pulsed-power lineage; smaller war chest
Private; laser inertial fusion leveraging NIF ignition physics - direct inertial-confinement rival with a different (laser) driver
Private; field-reversed configuration, more than $1.2B raised over two decades; cautionary tale on timelines
Private (Canada); magnetized target fusion, public funding difficulties and layoffs in 2025 - shows the downside scenario for capital-intensive fusion
Public fission microreactor developer - competes for the same data-center firm-power demand on a nearer timeline
Public SMR fission vendor - substitute product for firm clean baseload this decade