Analysis
Valar Atomics has raised a $1 billion Series B led by Sequoia Capital, tripling its valuation to $6 billion, according to Bloomberg. Founder Isaiah Taylor, a high school dropout who taught himself reactor engineering and nuclear physics before founding the company in his twenties, is building small modular reactors intended to be manufactured on a factory line and shipped to site, rather than engineered and poured in place the way conventional nuclear plants have been built for seventy years -- the single biggest cost and schedule driver in nuclear construction.
The round lands squarely inside the AI industry's power crunch. Data centers built for training and inference now draw enough electricity that hyperscalers are directly financing generation capacity rather than waiting on utilities: Pulse has covered Nvidia's $105 billion backing of an OpenAI data center in Ohio and Anthropic's new Theseus Infrastructure venture with Macquarie and GIC, both responses to the same underlying constraint -- interconnection queues in most U.S. grid regions now stretch multiple years, longer than most AI companies' compute roadmaps.
Valar's competitive set -- Oklo, Kairos Power, TerraPower and X-energy -- is racing toward the same factory-built, small-modular-reactor thesis, and none of them has yet delivered a commercially operating unit at scale; the entire category remains pre-revenue on its core product even as billions in venture and strategic capital pour in -- Oklo alone has raised capital via a SPAC listing on the strength of forward power-purchase agreements rather than delivered megawatts. Valar's differentiation, per its own pitch, is manufacturing-first design intended to compress the years-long, single-project construction timeline that has killed nuclear's cost competitiveness against gas and renewables for decades.
“Until then, the $6 billion mark is priced almost entirely on execution risk the market has not yet had a chance to test.”
A $6 billion valuation for a company with no operating reactor is a bet on regulatory approval timelines and manufacturing execution more than on any revenue Valar can show today -- nuclear projects live or die on Nuclear Regulatory Commission licensing schedules that have historically run years longer than sponsors project at the outset. The counterweight to the AI-power-crunch narrative driving this round is that even a successful factory-built reactor design still needs NRC sign-off project by project, and that regulatory bottleneck, not manufacturing cost, has been the more persistent constraint on U.S. nuclear buildout.
Sequoia's involvement is itself notable: the firm has historically stayed away from capital-intensive, decades-long infrastructure bets in favor of software's faster return cycles, and its willingness to lead a $1 billion nuclear round signals how far mainstream venture capital has moved toward funding the physical infrastructure AI now depends on -- power, chips, data centers -- rather than only the software layer sitting on top of it.
Whether Valar or any of its small-modular-reactor peers hits a first commercial online date before 2030 is the timeline against which every dollar of this round, and every AI data center counting on nuclear power to arrive on schedule, is ultimately being underwritten. Until then, the $6 billion mark is priced almost entirely on execution risk the market has not yet had a chance to test.