The rapid expansion of artificial intelligence is creating a new infrastructure challenge that extends well beyond processors, models and software. As chip performance accelerates and power densities rise, the ability to remove heat efficiently is becoming a defining constraint on future AI growth.
Against that backdrop, cooling specialist ZutaCore has announced a Series C funding round worth more than $100 million, backed by investors including Mitsubishi Electric, Carrier Ventures and Samsung Electronics through Samsung Ventures. The company said the investment will support international expansion and further research into cooling technologies designed for next-generation AI and high-performance computing infrastructure.
The announcement reflects a broader shift taking place across the data centre industry. As AI workloads become larger and more computationally intensive, traditional cooling approaches are coming under increasing pressure. The challenge is particularly acute as processors continue to demand higher power levels while operators seek to maximise computing density within existing facilities.
According to ZutaCore, its waterless two-phase direct-to-chip cooling technology is designed to support future AI and HPC processors exceeding 4,000 watts and to enable higher-density deployments at scale. The company said demand for its technology is being driven by growing investment in AI infrastructure worldwide.
Heat becomes the new bottleneck
The significance of cooling technology has grown alongside the rapid evolution of AI hardware. While much attention has focused on the development of increasingly powerful processors, industry discussions have increasingly shifted towards the practical realities of operating them.
ZutaCore said it has completed more than 75 deployments across the Americas, Europe and Asia, highlighting growing adoption of liquid cooling technologies in production environments. The company is also investing in technologies designed to integrate with existing air-cooled and single-phase cooling systems, reflecting the reality that many operators must modernise infrastructure incrementally rather than through wholesale replacement.
The funding will also support research into thermal management technologies aimed at addressing the changing requirements of future chip architectures and supporting megawatt-class deployments.
Infrastructure at AI scale
The company’s announcement underlines how infrastructure challenges are increasingly shaping the economics of AI deployment. As organisations race to deploy larger models and more powerful computing clusters, questions around power consumption, cooling capacity and facility design are becoming strategic concerns rather than purely operational ones.
To support larger-scale deployments, ZutaCore has established a 2MW end-of-row emulation platform in Israel. The facility is designed to replicate thermal and facility interactions at multi-megawatt scale without relying on production IT equipment, allowing operators to validate performance and integration requirements before deployment.
The company has also continued developing products aimed at AI infrastructure, including a cooling solution designed for the NVIDIA RTX PRO 6000 Blackwell Server Edition operating in standard enterprise and AI cloud server environments.
Alongside the funding announcement, ZutaCore confirmed the appointment of four senior executives across finance, people, research and development, and operations as it expands its international presence and prepares for further growth.
The significance of the announcement extends beyond a single company. It illustrates how the AI infrastructure race is increasingly being fought on multiple fronts. Advances in models and silicon remain important, but the ability to power, cool and operate those systems efficiently is becoming equally critical. As AI workloads continue to scale, technologies once regarded as supporting infrastructure are moving closer to the centre of the industry’s strategic agenda.
