Accelsius Targets AI Data Center Demand With NeuCool IR150 Rack-Level Cooling Solution

According to a recent LinkedIn post from Accelsius, the company is introducing the NeuCool IR150, described as a two-phase coolant distribution unit (CDU) integrated into a rack-level cooling solution for data centers. The post indicates that the IR150 combines a two-phase CDU, 42U of IT rack space, and liquid and vapor manifolds in a single 800mm-wide enclosure, with a stated capacity of up to 150 kW.

The LinkedIn content suggests the system is designed as a plug-and-play option for high-density, two-phase-enabled deployments, aiming to simplify installation and reduce technology cooling system complexity and cost. It also notes that the design is intended to confine cooling-related failure domains to a single rack, which could be relevant for operators seeking to limit downtime impact in AI and other mission-critical workloads.

A quote attributed in the post to CEO Josh Claman frames the IR150 as an evolution in data center infrastructure aligned with anticipated liquid cooling needs for next-generation AI supercomputers. This positioning, particularly the reference to “the world’s leading chipmaker” designing AI systems around liquid cooling, may indicate Accelsius is targeting the rapidly expanding AI and high-performance computing segment, where thermal management is increasingly a key constraint.

The post further mentions that the NeuCool IR150 will be showcased at NVIDIA GTC 2026 in San Jose, with availability for meetings and demonstrations via direct contact. For investors, visibility at a major AI-focused industry event and emphasis on high-density, rack-level liquid cooling could signal an effort by Accelsius to capture share in the emerging infrastructure layer supporting AI data centers, potentially enhancing its relevance in the broader data center and semiconductor cooling value chain.