Iceotope closed a $26 million Series B on May 14, co-led by Two Seas Capital and Barclays Climate Ventures, with existing backers Edinv, ABC Impact, Northern Gritstone, and British Patient Capital joining the round. DCD's coverage reports the UK company will direct the capital primarily at product development, engineering headcount, patent work, and ecosystem partnerships. Co-founder Simon Jesenko holds the CEO and CFO roles, which itself signals a company still operating lean.
The two products in market are KUL AI, a rack-scale system aimed at hyperscale and cloud deployments, and KUL BOX, a compact unit for individual servers and edge workloads. Both run Iceotope's single-phase precision immersion approach using a non-flammable dielectric fluid rather than water at the chip.
Twenty-six million dollars is a modest round against the recent comparison set. Accelsius raised $65 million in its Series B. Ecolab's acquisition of CoolIT cleared $5 billion. Iceotope holds one of the deeper patent positions in the sector, with a milestone of more than 200 granted patents in precision liquid cooling. The gap between that IP depth and a $26 million growth round is the actual signal here. Capital is flowing heavily into cold plate vendors with hyperscaler design wins. Immersion-adjacent architectures are raising smaller rounds at a slower cadence.
Jesenko framed the raise around readiness to scale at a moment of peak demand. SemiAnalysis projects liquid-cooled AI accelerator capacity growing from roughly 3 gigawatts to more than 40 gigawatts within two years. That is the market Iceotope is funding against. A 10x expansion in two years rewards whoever can manufacture and ship at volume, and a portfolio of patents does not cool a single GPU until it is in a product on a dock.
The structural challenge is not the technology. It is that direct-to-chip cold plate cooling has become the default specification for Blackwell and Rubin class racks, and immersion keeps losing that procurement decision on integration friction rather than thermal performance. Iceotope's sealed chassis design is a deliberate answer to that friction, since it avoids tank-scale fluid handling and fits standard rack mechanicals. The Series B reads as a bet that the answer is correct and that the company can prove it at production scale before the window closes.
For operators, the practical takeaway is supplier risk. The liquid cooling supply chain is consolidating around a small group of well-capitalized cold plate vendors. A $26 million round keeps Iceotope in the race but does not move it into the frontrunner tier on balance sheet alone. Anyone specifying precision immersion for a 2027 build should price in the financing trajectory, not just the datasheet, and ask Iceotope directly what production capacity this round actually funds.
Two markers will tell the story. The first is a named hyperscale or large colocation deployment of KUL AI at meaningful rack count, which would convert the patent narrative into a reference customer. The second is the next raise. If the follow-on is another sub-$50 million round, the funding gap between immersion and cold plate hardens into a structural disadvantage. If it is a nine-figure round tied to a design win, the precision immersion thesis is back in play. The $26 million does not settle that question. It buys the time to answer it.