Quaise Energy has raised $134 million in the first close of a Series B round led by Prelude Ventures, with Japanese energy majors JERA and Idemitsu joining as strategic investors. The round brings Quaise’s total funding to $230 million and is earmarked specifically for Project Obsidian, an attempt to build the world’s first commercial superhot geothermal power plant, sited south of Oregon’s Newberry Volcano.
The technology at the center of the bet traces back further than the company itself. Paul Woskov, a senior research engineer at MIT’s Plasma Science and Fusion Center, began experimenting in the early 2000s with gyrotrons, high-power millimeter-wave sources originally built for nuclear fusion research, and found in lab tests that the waves could vaporize granite by heating it to a plasma state, achieving drilling rates far beyond mechanical bits without the wear a physical drill bit suffers. By 2008 Woskov had proposed applying the same millimeter-wave approach specifically to reach deep geothermal resources, addressing exactly the depth limitation conventional rotary drilling runs into past roughly 10 kilometers. Quaise itself wasn’t founded until a decade later, in 2018, by Carlos Araque, now CEO, and Matt Houde, after Araque was introduced to Woskov’s research in 2017 and set out to commercialize it.
Quaise’s funding history shows the same slow-then-fast pattern common to deep-tech companies waiting for their underlying science to become commercially credible. The company raised a $6 million seed round in June 2020 led by The Engine, the MIT-affiliated venture firm, and Collaborative Fund, with Vinod Khosla as its first backer. A $21 million Series A1 followed in 2024, led by Prelude Ventures and Safar Partners with Mitsubishi Corporation and Standard Investments joining as new investors, mirroring this new round’s mix of financial and strategic energy capital. This Series B, at $134 million, is more than six times the size of that Series A1, a jump that reflects both AI-driven power demand pulling forward investor interest in any credible new firm-power source and Quaise’s own technical progress moving the drilling technology from lab demonstration toward an actual field deployment plan.
Conventional geothermal drilling reaches rock hot enough to be useful but well short of the 300 to 500 degrees Celsius range where superhot geothermal becomes viable at scale, and drilling mechanically past a certain depth becomes prohibitively slow and expensive as the rock gets harder and hotter. Quaise’s approach swaps the drill bit for that millimeter-wave energy beam past that threshold, targeting drilling depths beyond five kilometers that would be impractical with conventional equipment.
The commercial logic behind Project Obsidian is unusually concrete for an early-stage energy technology company: Quaise has already signed a hyperscaler customer for the project’s first 50 megawatts, meaning a major cloud or AI-compute company has committed to buying power from a plant that doesn’t exist yet, built on drilling technology that has never operated at this depth commercially. That’s a significant bet by the hyperscaler customer, but it reflects how urgently AI-scale data center operators are chasing any credible source of firm, always-on clean power, geothermal’s key advantage over solar and wind, rather than waiting for conventional nuclear or grid expansion timelines that stretch a decade or more.
JERA and Idemitsu’s participation is worth noting specifically. Both are established Japanese energy conglomerates, not typical venture investors, and their involvement signals that traditional energy incumbents are treating next-generation geothermal drilling as a strategic technology worth a direct equity stake in, rather than something to watch from the sidelines and potentially license or acquire later. That’s a pattern increasingly visible across 2026’s energy-for-AI funding wave, corporate energy players moving earlier and more directly into venture-stage climate and power technology than they typically have in prior cycles, driven by the sheer scale of new electricity demand AI data centers are creating globally.
Quaise’s raise sits inside a broader surge of capital flowing specifically into power infrastructure built to serve AI-scale compute demand, a category that has produced some of 2026’s largest individual funding rounds across energy, grid, and generation technology. The throughline across all of these deals is the same: AI compute growth has outpaced realistic near-term grid capacity in multiple major markets, and investors are betting heavily on whichever combination of nuclear, geothermal, and grid technology can bring new firm power online fastest, with geothermal’s pitch being that, unlike nuclear, it doesn’t require multi-decade regulatory approval cycles once the drilling technology itself is proven.
The Philippine angle here is more structurally relevant than for most global energy-technology stories, and the numbers back that up more concretely than a general geological similarity would suggest. The Philippines is the world’s third-largest geothermal power producer, behind only the United States and Indonesia, with roughly 1,984 megawatts of installed capacity supplying more than 17 percent of the country’s electricity mix, a far larger share of the national grid than geothermal provides almost anywhere else on earth. Energy Development Corporation, a First Gen subsidiary, is the world’s largest vertically integrated geothermal power company and alone accounts for about 1,484 megawatts, roughly 61 percent of the country’s total geothermal capacity, having recently brought its 22-megawatt Tanawon plant in Sorsogon online. Conventional Philippine geothermal technology doesn’t reach the superhot depths Quaise is targeting, but the country’s existing geological advantage, institutional operating experience, and geothermal engineering talent base built up over decades through EDC and its predecessors are exactly the kind of foundation that could make next-generation superhot drilling technology, once proven commercially in Oregon, a plausible eventual fit for Philippine deployment, particularly as the country’s own data-center and AI-infrastructure ambitions around the Clark Freeport Zone hub create the same kind of firm-power demand driving Quaise’s hyperscaler customer today.
That fit is a long-horizon possibility, not a near-term opportunity. Quaise’s technology hasn’t yet proven itself at commercial scale even in its first US deployment, and any Philippine application would require years of technology transfer, capital, and regulatory groundwork after Project Obsidian itself succeeds or fails. But for a country whose energy planners already think in geothermal terms by default, and whose largest geothermal operator already runs more installed capacity than most countries have in total, it’s a technology worth tracking closely rather than treating as a purely American story.
Share this article