Accelsius introduced its NeuCool IR150, a fully integrated rack-level cooling system designed to deliver up to 150kW of thermal capacity within a standard 800mm enclosure. The system combines a two-phase Coolant Distribution Unit (CDU), 42U of IT rack space, and built-in liquid and vapor manifolds, targeting rapid deployment across hyperscale, neocloud, and enterprise environments.
The IR150 reflects a shift toward modular, plug-and-play liquid cooling architectures that can move through conventional IT procurement and deployment channels. Accelsius positions the system for both centralized AI data centers and distributed edge environments, particularly as smaller AI models and inference workloads proliferate. Alongside the product launch, the company introduced its NeuCool HyperStart program, aimed at helping operators validate two-phase direct-to-chip cooling designs and integrate them into future data center reference architectures.
The company frames its two-phase approach as an alternative to traditional single-phase liquid cooling systems, which rely on water-based cooling loops. Accelsius uses a non-conductive dielectric refrigerant designed to avoid water exposure at the rack level, reducing risks associated with leaks and corrosion. The company cites third-party and internal analysis indicating potential reductions in cooling energy consumption of up to 90 percent versus air cooling, along with 35–44 percent annual OpEx savings and 8–17 percent five-year TCO improvements compared to single-phase direct-to-chip systems.
- NeuCool IR150 delivers up to 150kW per rack in a fully integrated 42U enclosure
- Combines CDU, rack infrastructure, and liquid/vapor manifolds in a single unit
- Designed for plug-and-play deployment through standard IT channels
- Targets hyperscale, neocloud, enterprise, and edge AI deployments
- Uses dielectric, non-conductive refrigerant with no water inside the rack
- HyperStart program supports validation and deployment of two-phase cooling designs
- Complements existing NeuCool MR250 (250kW+) and Thermal Simulation Rack platforms
“Hyperscalers and neoclouds are under enormous pressure to deliver AI capacity faster than ever, and that urgency is understandable,” said Josh Claman, CEO of Accelsius. “With the general availability of the IR150 and the launch of NeuCool HyperStart, we’re giving operators a practical path to bring two-phase, direct-to-chip cooling into their reference designs today.”
🌐 Analysis: The introduction of a rack-integrated two-phase system aligns with a broader industry shift toward higher-density, liquid-cooled AI infrastructure, where rack power levels are moving beyond 100kW. Vendors across the ecosystem—including cooling specialists and OEM server providers—are racing to productize modular liquid cooling systems that can be deployed without custom engineering. Accelsius’ emphasis on waterless operation and standardized form factors targets a key friction point in enterprise adoption.
🌐 Analysis: The launch also comes as data center developers face increasing regulatory and community scrutiny over water and power usage. Competing approaches—such as single-phase direct-to-chip cooling, rear-door heat exchangers, and immersion cooling—are evolving in parallel, with hyperscalers often pursuing hybrid strategies. Accelsius’ HyperStart program suggests growing demand for validated reference designs as operators standardize cooling architectures for next-generation AI factories.
