Co-Authored by:
Mason Wolak, VP, Capacitor Development at Peak Nano
& Caroline Sorrick, Founder and CEO of E&P Technologies
Fusion’s promise has always been easy to describe yet hard to engineer: harness the power of the sun and put it on the grid. Over the last few years, fusion companies have started designing inertial confinement fusion (ICF) architectures with commercialization in mind, but using traditional pulse power systems to provide the enormous energy required to initiate fusion reactions presents significant technical and financial roadblocks.
To address this gap in performance and drive down both cost and footprint, a new class of impedance‑matched generator (IMG) pulse power systems is emerging. Whereas legacy Marx generators required huge banks of capacitors and were largely limited to single-shot experiments, IMG-enabled fusion can repeatedly generate equivalent energy outputs in a fraction of the size. These machines direct energy into a target as efficiently as possible, delivering megajoules of energy in nanoseconds, repeatedly and reliably.
This shift from one-shot spectacle to workhorse repetition changes the engineering calculus for every component downstream. Pulsed-power hardware, particularly high-voltage capacitors and the dielectric films inside them, now must survive not one extreme event, but millions. IMG-enabled fusion pushes these components into new regimes of electrical, thermal, and mechanical stress for which traditional materials weren't designed.
From “Science Machines” to Modular Power Infrastructure
Historically, pulsed‑power fusion systems were built around a single goal: to achieve the highest possible peak power in a single experimental shot. Classic Marx generators, first invented in 1924, operated at extreme voltages but required multiple pulse‑compression stages, water tanks, and pulse‑forming lines. While delivering impressive physics, they were expensive to build, complex to maintain, and impossible to replicate at power‑plant scale.
The impedance‑matched generator is the first major innovation in Marx generator design in more than 90 years. Instead of one monolithic machine, these systems use large numbers of standardized modules whose electrical impedance is matched directly to the load. That design choice has three concrete consequences:
- Pulses can be formed at the load itself, eliminating bulky compression lines
- Energy reflected back into the capacitors can be minimized
- Cost per joule delivered drops sharply as modules are replicated
Leading fusion development programs now envision systems built from hundreds of IMG “power blocks.” Pacific Fusion’s demonstration system, for example, uses 156 identical modules to generate high peak currents and fusion output, with the aim of achieving net facility gain by 2030.
Advanced pulsed power architectures like the IMG open a path to mass manufacturing, but current technology gaps must be bridged to achieve the durability, high repetition rates, and affordability necessary for commercialization.
Redesigning Capacitors for Fusion’s Power and Speed
For capacitor designers, the IMG-enabled fusion revolution changes the specs. These architectures require pulse‑power stacks to store and discharge megajoules of energy with nanosecond-scale rise times, sustaining the repetition rates needed for power-plant duty cycles. Hardware must survive millions to billions of shots in high‑temperature environments near power electronics, switches, and radiation sources.
Traditional film capacitors built with biaxially oriented polypropylene (BOPP) weren’t designed for this. They require complex cooling, take up too much volume, and are difficult to manufacture in the quantities IMG-enabled fusion plants will need… tens of thousands per grid-integrated facility.
In other words, IMG physics is only half the challenge. The other half is making the capacitors behind it small, robust, and manufacturable, so they can be populated in hundreds of synchronized modules.
How Nanolayer Dielectric Films Help
Peak Nano has been working on this problem from the materials side. Its NanoPlex™ HDC capacitor films were developed with national labs to deliver a higher dielectric constant, improved resistance to high-field breakdown, higher-temperature operation, and better volumetric efficiency than conventional films. They are now being refined as a fusion-centric solution through collaboration with fusion programs.
Here’s how nanolayering works: Multiple polymers are arranged into tens to thousands of ultra‑thin layers, each tunable for specific electrical and thermal performance. This creates a composite scaffold that outperforms any single material. NanoPlex HDC delivers dielectric constants in the 3.7–4.8 range today (with a roadmap toward 8x BOPP energy density) and operating temperatures of 105°C and beyond, resulting in significantly smaller capacitor volume.
Because Peak uses off‑the‑shelf resins and runs them on industry‑standard extrusion lines, the incremental materials cost is less than 10% above that of traditional films. For IMG, these performance gains translate into denser power blocks, smaller banks, and less cooling infrastructure, making it possible to stack hundreds of modules into a commercial ICF driver while creating direct savings in footprints, bill of materials costs, and system efficiencies.
Turning Film Innovation into Pulsed‑Power Hardware
On the hardware side, E&P Technologies was founded to address gaps that have long been felt by fusion developers: long lead times, limited suppliers, and capacitors that aren’t optimized for fusion’s extreme duty cycles.
E&P designs and manufactures advanced high‑voltage capacitors (5 kV up to 250 kV) for fusion pulsed power, defense, and other extreme‑energy environments. Several aspects of their approach map directly onto IMG-enabled fusion requirements:
- Capacitor “bricks” are designed as line‑replaceable units that can be stacked in series or parallel and integrated into IMG power blocks without bespoke redesign.
- An AI-assisted design platform explores geometries and winding configurations that hit electrical targets while optimizing manufacturability and cooling.
- Embedded sensors and telemetry can feed health data into plant-level dashboards, enabling predictive rather than time-based maintenance.
- U.S.-based pod-style manufacturing uses automated winding and testing cells to deliver high throughput with consistent quality, strengthening domestic supply chains.
Peak Nano and E&P Technologies have recently launched a joint development program that takes NanoPlex HDC films and packages them into compact, modular capacitor designs optimized for IMG-enabled fusion’s high‑shot‑count environment. The goal isn’t just higher performance per device, but to create a platform that can scale for future IMG power plants.
Designing for “Megajoules in Nanoseconds”
IMG-enabled fusion is still in its early chapters, but the direction of travel is clear. The film capacitor is becoming a strategic component. Executing on IMG’s powerful specifications requires us to rethink capacitor design from all angles, from material selection to film winding, architecture, and plant‑level maintenance.
A few principles are emerging from our work:
- Start with IMG modularity in mind. Capacitors should be standard, line‑replaceable units aligned with module footprints and connector schemes, not one‑off designs retrofitted later.
- Design for diagnostics, not just ratings. Embedded sensing and data capture are essential if operators are going to trust multi‑megajoule systems to run at high repetition rates for decades.
- Co‑develop materials and hardware. Concurrent tuning of film properties and capacitor device design and fabrication dramatically shortens iteration and qualification timelines.
- Build domestic capacity early. When thousands of IMG plants are deployed globally, the first countries to ship hardware will define the supply chain for decades.
- Engage directly with end users. Conduct a structured needs analysis with fusion developers to understand real‑world capacitor challenges and solve the problems that matter most.
By aligning materials science with pulsed‑power engineering now, we can ensure that when IMG systems reach the grid, their megajoules arrive exactly when and where they’re needed.
Watch our recent webcast to learn more about how IMG-enabled fusion is reshaping hardware demands.