Accelsius, which offers a proprietary, two-phase, direct-to-chip liquid cooling system to support HPC and rack density for data center and edge operators, successfully raised $24 million pursuant to a Series A funding round. The round initiated in early 2024.
“This investment is a testament to our team’s intense focus on operational innovation and technological ingenuity,” said Josh Claman, Accelsius’ CEO. “We believe that with AI’s rapid growth, we all must aim to make the surrounding infrastructure as efficient as possible. Liquid cooling is the future of data center cooling, and Accelsius is positioned to lead that future.”
Accelsius will use the funds to accelerate plans for an international presence, continuing to grow U.S.-based shipments and revenue while also expanding its professional teams across engineering, R&D, sales, marketing, manufacturing and operations.
Funding for this round has been provided from and privately raised by Innventure, who founded Accelsius in June 2022. Innventure is a technology commercialization platform that founds, funds and supports the launch and scale of its companies with a focus on transformative technology solutions.
“From its beginning, Accelsius has exhibited all the qualities of a market disruptor,” said Bill Haskell, CEO of Innventure. “We are proud to see all the great progress that Josh and his team have made to produce critical infrastructure required to support the massive growth in data centers and related markets. To go from founding to revenue generation in less than three years is truly impressive, and we couldn’t be more excited for Accelsius’ future.”
Accelsius’ NeuCool system addresses the cooling challenges of modern data centers by delivering gains in efficiency and sustainability. Compared to conventional cooling methods, NeuCool offers an estimated 50 percent reduction in energy use and 10X rack power density improvement, all while eliminating water consumption. Tested to handle thermal loads exceeding 2,200 watts per server chip (CPU, GPU), NeuCool is optimally suited for the high-powered servers that drive AI, ML, HPC and other computationally intensive workloads that are prevalent in today’s data centers.
