ASUS has unveiled its latest advancements in liquid-cooling technology aimed at enhancing the performance of next-generation AI and high-performance computing (HPC) data centers. The company introduced its ASUS Optimized Liquid-Cooling Solutions and a strategic partner framework designed to tackle increasing thermal, power, and density challenges faced by modern computing systems.
As workloads from AI and HPC push the limits of traditional air cooling, these optimized solutions are vital for maintaining operational efficiency. The liquid-cooling systems are specifically tailored to support the upcoming NVIDIA Vera Rubin NVL72 architecture. By providing effective heat dissipation for high-performance CPUs, GPUs, and densely packed accelerator racks, ASUS aims to significantly reduce energy consumption and improve power usage effectiveness (PUE), while also optimizing total cost of ownership (TCO).
Comprehensive Cooling Solutions and Strategic Partnerships
The ASUS portfolio includes several innovative cooling configurations, such as direct-to-chip (D2C) cooling, in-row cooling with chiller distribution units (CDU), and hybrid systems. These solutions are developed in collaboration with leading infrastructure companies, including Schneider and Vertiv, as well as precision component manufacturers like Auras Technology and Cooler Master.
ASUS has established a framework of strategic partnerships that enables it to deliver tailored cooling solutions, ensuring optimal stability and performance for large-scale AI deployments. The company boasts an impressive record in the industry, holding 2,156 No. 1 records in SPEC CPU benchmarks and 248 No. 1 results in MLPerf, underscoring its leadership in compute density and AI performance.
Notable Deployment in Taiwan
A standout example of ASUS’s liquid-cooling technology is its recent project with the National Center for High-performance Computing (NCHC) in Taiwan. This installation features a dual-compute architecture, integrating the NVIDIA HGX H200 cluster and the NVIDIA GB200 NVL72 system. It marks Taiwan’s first deployment of a fully liquid-cooled AI supercomputer using this architecture.
Designed entirely by ASUS, the system employs direct liquid cooling (DLC) technology to achieve an exceptional PUE of just 1.18. This deployment exemplifies the seamless integration of high performance with sustainable design, highlighting ASUS’s expertise in thermal and energy management within expansive AI infrastructures.
ASUS will showcase its innovations at the NVIDIA GTC 2026, where it will serve as a Diamond Sponsor from March 16–19, 2026, in San Jose, USA. Under the theme “Trusted AI, Total Flexibility,” ASUS plans to collaborate with NVIDIA and other global infrastructure leaders to present a robust, next-generation liquid-cooling ecosystem.
Availability and Future Prospects
ASUS servers equipped with these cutting-edge cooling solutions are available worldwide. Interested parties should contact their local ASUS representatives for pricing and additional information.
With these advancements, ASUS continues to position itself at the forefront of AI infrastructure, offering solutions that not only meet the demanding requirements of today’s data centers but also pave the way for future innovations in the industry.
