A $10M pre-seed puts California's first new reactor plan on a barge at Long Beach

Start with what the money is a fraction of. If nuclear development costs routinely run into the billions, as Crypto Briefing's account of the round notes ^[17], then $10 million ^[1] covers one percent of a billion-dollar program and half a percent against two billion ^[18], which makes this a round that buys a position rather than a plant. Divided the other way, against a 10 MWe design ^[3], the equity raised so far works out to a dollar per nameplate watt ^[21].

What it does buy is sequencing. The port and the Maritime Administration signed a non-binding cooperation agreement in July 2026, the first of its kind for a US port ^[6], roughly two months ^[19] after MARAD issued a request for information on US-built commercial SMRs in May 2026, the process the partnership grew out of ^[8]. The described work is standards work, jointly with the Coast Guard and the Department of Energy, on safety and operations for floating nuclear power ^[7]. Meanwhile the company is in engineering and regulatory preparation with no design application filed at the Nuclear Regulatory Commission ^[13], and no nuclear fuel or material on site ^[14]. The hardware is real enough: a 185-foot tank barge acquired in June 2026 and delivered along with an electric test reactor pressure vessel ^[15][2], sitting in a state that effectively outlawed new nuclear plants about half a century ago ^[5].

The load arithmetic is worth doing before anyone reaches for the data-center story. Ten megawatts electric across the roughly 15,000 homes the design is said to serve ^[3] is about 667 watts per home on a continuous-average basis ^[20], which is the scale of an industrial hookup, and indeed the first named market is shoreside power for ports and terminals in place of diesel generators, with vessels and data centers held for later ^[9]. Long Beach is the second-busiest container port in the country, with zero-emissions goals running to 2050 alongside plans for higher container throughput ^[12], so berth-side supply that displaces diesel has an obvious buyer at a price nobody has published.

This is probably wrong, but the durable asset here looks like the standards file rather than the reactor, since whoever helps write the first Coast Guard and DOE rulebook for a floating reactor shapes everyone else's application. It could go otherwise in at least three ways. The permitting clears and the economics do not, which is the NuScale case, where the furthest-along US developer saw its first planned deployment cancelled in 2023 after cost overruns ^[16]. The agreement quietly expires, non-binding being a term that cuts both ways. Or, or rather the more interesting version, a better-capitalised party buys the position, because floating reactors are not an engineering unknown: the US Navy has run reactors at sea for decades ^[10] and Russia put a barge-mounted plant into service for a remote Arctic town in 2020 ^[11]. What would prove the thesis wrong is a docketed NRC application with a priced port offtake behind it before any standard appears, which would mean the licence, not the rulebook, was the scarce thing all along.