The a16z Podcast
The a16z Podcast

How Radiant and Heron Are Rethinking Power Generation and Delivery

a16z general partners Erin Price-Wright and Erik Torenberg speak with Doug Bernauer, founder and CEO of Radiant, and Drew Baglino, founder and CEO of Heron, about rebuilding American energy infrastructure. They discuss portable micro nuclear reactors, solid state power electronics, why delivery rath

Featured Speakers

a16z Host

Topics Discussed

Episode Summary

Executive Summary: The episode argues that U.S. energy constraints are increasingly about delivery, not generation, as AI, data centers, electrification, and reindustrialization drive demand. It spotlights two edge-first solutions: Radiant’s trailer-sized microreactors and Heron’s solid-state power electronics, both aimed at making power more modular, resilient, and manufacturable.

Main Topics: The grid’s bottleneck is delivery, not new generation (Priority: 5/5): The speakers frame the core energy problem as an aging, centralized transmission and distribution system that cannot move power fast enough to meet new demand from data centers, industry, and electrification. Radiant’s portable microreactor strategy (Priority: 5/5): Doug Bernauer explains Radiant’s factory-built, trailer-sized nuclear reactor designed for off-grid or edge use cases like military bases, disaster relief, and remote industrial sites. Heron’s solid-state transformer and grid modernization (Priority: 5/5): Drew Beglino describes Heron’s bidirectional power electronics that convert and manage power more efficiently than traditional transformer infrastructure, enabling more flexible DC/AC systems. AI and reindustrialization as a forcing function (Priority: 4/5): The conversation argues that AI data centers, defense, and reshoring are exposing the limits of the current grid and accelerating investment in resilient, software-defined power systems. Manufacturing, modularity, and factory production (Priority: 4/5): Both founders emphasize building energy hardware like products—standardized, modular, and mass-manufacturable—rather than as bespoke megaprojects assembled on site. Nuclear fuel, waste, and regulatory bottlenecks (Priority: 4/5): The discussion highlights that nuclear’s biggest barriers are not physics but fuel access, enrichment, waste storage, and permitting infrastructure. DC microgrids and convergence of new loads (Priority: 3/5): The speakers note that batteries, solar, compute, and microreactors are naturally DC-oriented, suggesting a future where local DC microgrids integrate multiple power sources more cleanly.

Key Arguments: New power generation is not the main bottleneck; the real constraint is getting electricity delivered through an outdated grid. Decades of efficiency gains masked grid underinvestment, but demand from AI, transport, and industrial electrification is now outpacing those gains. Data centers are becoming large, spiky loads that can destabilize the grid unless paired with better controls, storage, and power electronics. Data centers can also lower average electricity costs by increasing utilization of the grid and spreading fixed costs over more kilowatt-hours. Radiant’s microreactor is positioned as an off-grid alternative to diesel generators for customers who need resilient power but not gigawatt-scale plants. Heron’s solid-state transformer replaces bulky, oil-based, line-frequency transformers with high-frequency semiconductor-based systems that are smaller, faster, and more controllable. Both companies are pursuing factory-based manufacturing because it improves quality, cost, speed, and repeatability compared with field-built infrastructure. The nuclear industry is still early; meaningful progress depends on fuel supply, enrichment, waste repositories, and a more mature regulatory ecosystem. A future energy system will likely be hybrid: solar, batteries, nuclear, and power electronics working together in modular architectures. The grid should become more software-defined, decentralized, and bidirectional to support modern loads and edge generation.

Data Points: Edison’s first grid deployment: 85 customers across one square mile - Used to illustrate the original centralized model of electricity delivery in 1882 lower Manhattan. U.S. electricity demand trend: Rising for the first time in decades - Framed as a major shift driven by data centers, electrified transport, and reshoring. Grid growth need: 3 to 5x - Estimate mentioned for how much the electricity grid may need to expand. Radiant reactor size: 1 megawatt - The company’s trailer-sized microreactor product. Radiant deployment time: 48 hours - Time from arrival on site to power-on for the trailer-mounted reactor. Radiant runtime: 5 years - How long the reactor can operate before replacement or servicing. Diesel equivalence: 2 million gallons - Approximate diesel-equivalent energy output of the reactor over its operating life. Radiant factory site: 80 acres - Tennessee site signed for reactor production and fuel handling. Radiant production target: 1 reactor per week - Stated goal for output from the Tennessee facility. Heron product power rating: 5 megawatts - Heron Link, the first product, is a bidirectional solid-state transformer. Heron voltage range in: 800 to 1500 volts DC - Input range for Heron Link. Heron voltage range out: 34,000 volts AC - Output voltage for Heron Link, aligned with sub-transmission/distribution use cases. Heron module count: 30 modules - The 5 MW system contains 30 165-kW modules. Heron module size: 165 kilowatts each - Internal modular building blocks of the Heron Link system. Heron factory target: 40 gigawatts per year - Planned factory scale discussed as an optimal manufacturing cadence. Tack time: ~60 seconds - Referenced as the manufacturing cadence that maximizes capital efficiency. EV power semiconductor demand: 3 terawatts last year - Used to show the scale of power electronics supply chains already supporting electrification. U.S. peak grid power: Less than 1 terawatt - Compared against semiconductor demand to show the scale of the opportunity. Data center market size for DC products: ~500 gigawatts - Heron’s DC-focused market opportunity across data centers, solar, and batteries. Data center outage example: 2 gigawatts turned off instantaneously - Referenced as a recent event in Virginia illustrating grid instability concerns. Hong Kong diesel price: $10/gallon - Example of high-cost edge markets where microreactors can compete with diesel. Northern Europe diesel price: $7–$9/gallon - Another example of markets where off-grid nuclear can be economically attractive. U.S. nuclear waste repository example: Waste Isolation Pilot Plant in New Mexico - Cited as a successful defense-waste storage model that contrasts with the lack of a civilian repository.

Pivotal Quotes: "The grid is breaking." — Doug Bernauer: Opening framing of the episode’s central thesis about transmission and delivery constraints. "New power is not the problem. Delivery is the problem." — Narration / transcript framing: Summarizes the episode’s core argument about why generation alone won’t solve today’s energy challenge. "Electric power is civilization." — Doug Bernauer: Used to emphasize the foundational role of electricity and the importance of re-architecting the grid.

Implications: Energy infrastructure is shifting toward modular, software-defined, factory-built systems. Expect more edge power, DC microgrids, and hybrid architectures as AI and industry force faster grid modernization.

🔓 Sign Up for Unlimited Episode Search

About The a16z Podcast

The a16z Podcast discusses tech and culture trends, news, and the future – especially as ‘software eats the world’. It features industry experts, business leaders, and other interesting thinkers and voices from around the world. This podcast is produced by Andreessen Horowitz (aka “a16z”), a Silicon Valley-based venture capital firm. Multiple episodes are released every week; visit a16z.com for more details and to sign up for our newsletters and other content as well!

View all episodes from The a16z Podcast