As artificial intelligence workloads continue to push data center power densities to new levels, liquid cooling has become a critical component of next-generation AI infrastructure. Nautilus Data Technologies has announced that its EcoCore® Facility Cooling Distribution Unit (FCD) has achieved NVIDIA-Validated AI Factory Infrastructure status, becoming the first large-scale liquid-to-liquid cooling distribution unit validated under NVIDIA’s AI factory infrastructure program.
The rapid expansion of generative AI is driving unprecedented demand for high-density computing infrastructure, forcing hyperscale operators to rethink how next-generation GPU clusters are cooled. Nautilus Data Technologies has reached a new milestone in that evolving market after its EcoCore® Facility Cooling Distribution Unit (FCD) received recognition as NVIDIA-Validated AI Factory Infrastructure, validating its suitability for large-scale AI factory deployments.
The certification places the EcoCore FCD on NVIDIA’s DSX Infrastructure Marketplace, where infrastructure products are assessed against NVIDIA’s functional requirements for AI data centers. According to Nautilus, the system is the highest-capacity liquid-to-liquid cooling distribution unit currently listed on the marketplace and the first to achieve validation within the Large-Scale Cooling Distribution Unit category.
The announcement underscores the growing importance of liquid cooling as AI infrastructure evolves beyond the thermal limits of conventional air-cooled environments. Modern GPU clusters used for training and inference increasingly require advanced thermal management systems capable of supporting multi-megawatt deployments while maintaining energy efficiency and operational resilience.
Cooling Infrastructure Becomes a Strategic AI Asset
AI factories—large-scale data centers optimized for accelerated computing—place significantly higher thermal demands on infrastructure than traditional enterprise facilities. High-performance GPUs from vendors such as NVIDIA generate dense heat loads that require precision cooling to maintain performance, reliability, and operational efficiency.
According to Nautilus, its EcoCore FCD successfully completed validation against more than 100 technical requirements defined for NVIDIA AI factory deployments.
Independent testing reported cooling performance exceeding 4.1 megawatts in N-mode dual-pump operation and 3.2 megawatts in N+1 redundant operation. The company also reported an approach temperature difference (ATD) of 2°C, a metric used to measure heat transfer efficiency between cooling loops. Lower ATD values generally indicate more effective thermal exchange, which can contribute to improved energy efficiency in high-density AI environments.
By comparison, Nautilus noted that competing systems have reported higher ATD measurements, although comparative performance may vary depending on deployment architecture and operating conditions.
Designed for Modern AI Data Centers
Beyond raw cooling capacity, the EcoCore platform has been engineered to accommodate diverse deployment environments.
The cooling system supports multiple water sources—including chilled water, process water, seawater, river water, and treated wastewater—providing operators with greater flexibility when designing facilities in regions with varying infrastructure constraints.
Nautilus also emphasizes maintainability as a key design consideration. The company says the platform can be serviced and upgraded without requiring complete replacement and integrates with its Air Detection Removal System, a technology intended to improve liquid cooling reliability by removing entrained air from circulating coolant.
According to the company, the EcoCore platform has accumulated more than 650,000 operational unit-hours across three product generations and is available globally with delivery timelines of approximately 15 weeks.
Why NVIDIA Validation Matters
NVIDIA’s validation program has become an increasingly influential benchmark within the AI infrastructure ecosystem.
As enterprises, cloud providers, and hyperscale operators invest billions of dollars in AI-ready facilities, infrastructure components that meet NVIDIA’s deployment requirements can simplify procurement decisions and reduce integration risk.
The DSX Infrastructure Marketplace serves as a curated ecosystem of validated power, networking, cooling, and infrastructure technologies designed to support NVIDIA accelerated computing platforms. Inclusion signals that products have demonstrated compatibility with the operational requirements of modern AI factories, although overall deployment performance ultimately depends on broader system architecture.
For infrastructure vendors, NVIDIA validation is becoming an important differentiator as organizations standardize technology stacks around accelerated computing environments.
Market Landscape
Demand for AI infrastructure continues to accelerate alongside adoption of large language models and generative AI applications. According to IDC, worldwide spending on AI infrastructure is expected to grow at double-digit annual rates as enterprises and cloud providers expand accelerated computing capacity. Meanwhile, Gartner projects sustained investment in AI-ready data centers as organizations modernize facilities to support increasingly power-intensive GPU deployments.
This trend has intensified competition among infrastructure providers including Vertiv, Schneider Electric, CoolIT Systems, Supermicro, Dell Technologies, and NVIDIA, all of which are investing heavily in liquid cooling technologies, power distribution, and AI data center architecture.
Within this rapidly evolving ecosystem, Nautilus is positioning EcoCore as infrastructure built specifically for high-density AI environments rather than retrofitted enterprise cooling applications. As GPU thermal requirements continue to rise, liquid cooling is expected to become a foundational element of next-generation AI factories, making validated infrastructure increasingly important for organizations planning long-term AI capacity expansion.
Market Landscape
The AI infrastructure market is entering a new phase in which cooling technology is becoming as strategically important as compute hardware itself. As accelerator power consumption continues to increase, validated liquid cooling platforms capable of supporting multi-megawatt GPU clusters are emerging as critical building blocks for hyperscale AI data centers.
Top Insights
- Nautilus earned NVIDIA-Validated AI Factory Infrastructure recognition, positioning its EcoCore Facility Cooling Distribution Unit for deployment in next-generation AI data centers and hyperscale GPU environments.
- The EcoCore platform demonstrated cooling capacity exceeding 4.1 megawatts during independent validation, highlighting growing demand for liquid cooling systems capable of supporting increasingly dense AI workloads.
- Inclusion in NVIDIA’s DSX Infrastructure Marketplace strengthens EcoCore’s visibility among enterprises and hyperscale operators evaluating AI-ready power and cooling infrastructure.
- Flexible water-source compatibility and in-place serviceability reflect increasing emphasis on sustainable, resilient, and scalable cooling architectures for accelerated computing facilities.
- The announcement reinforces how thermal management is becoming a strategic component of enterprise AI infrastructure alongside GPUs, networking, and power distribution technologies.
Power Tomorrow’s Intelligence — Build It with TechEdgeAI
