Nuclear-inspired cooling that turns wasted power into more AI compute
Ferveret is developing Adaptive Phase Cooling, a next-generation thermal solution for AI data centers inspired by heat transfer principles used in nuclear reactor systems. Our rack-integrated design delivers 35% more usable compute from the same power envelope, with a PUE of 1.03 (vs. the 1.2 industry standard), zero water consumption, and 45°C inlet temperatures that eliminate the need for chillers. Compatible with NVIDIA MGX and HGX architectures, Ferveret's solution uses a proprietary dielectric fluid that is not part of any regulated PFAS category, giving AI infrastructure operators a scalable, water-free path to higher compute density without redesigning their facilities.
Total raised
$2.2M
Last stage
Seed
Investors
Matteo Bucci
Matteo is the co-founder and CTO of Ferveret as well as a faculty member in Nuclear Science and Engineering at MIT. Prior to Ferveret, he spent nearly a decade as a Research Scientist at CEA Paris. Fields of expertise include heat transfer, fluid dynamics, and nuclear reactor design. Outside of his core work, Matteo is also a nuclear energy consultant and editor of Applied Thermal Engineering.
LinkedInReza Azizian
MIT alumnus and former postdoc with 10+ years in the electronic cooling industry. Previously Senior Packaging Thermal Engineer at NVIDIA, Senior Design Thermal Engineer at Cruise, and designed thermal solutions for Microsoft HoloLens.
LinkedInNo applications, no recruiter spam. Just the intro.
A few questions to make sure this role is the right shape for you. Two minutes.
I write the intro, send it to the founder, and handle the back-and-forth.
If they’re a yes, I book the chat. You show up — that’s the whole job-hunt.