MSI Unveils Cutting-Edge Liquid-Cooled AI Solutions at COMPUTEX 2026

During COMPUTEX 2026, hosted in Taipei, MSI showcased an impressive array of AI and data center infrastructure solutions, prominently featuring innovative liquid-cooled technology. This portfolio is tailored to meet the growing demands for enhanced computational capacity while ensuring optimal thermal management in modern data centers.

Liquid-Cooled and Standard Rack Architectures

Central to MSI's offerings are the ORv3 liquid-cooled rack architecture and traditional air-cooled options, designed to cater to various deployment needs. The ORv3 architecture excels in high-density AI and cloud applications, facilitating up to 100kW deployments through an integrated Liquid-to-Liquid Coolant Distribution Unit (L2L CDU). Notably, this 21” 44OU setup incorporates 28 Open Compute multi-node systems, optimizing compute density and efficiency.

Meanwhile, the 19” 48RU EIA Air-Cooled Rack Architecture offers compatibility for established enterprise environments, supporting configurations with up to 16 2U2N multi-node systems. With AMD EPYC™ 9005 and Intel® Xeon® 6 options, users can choose based on performance requirements.

NVIDIA MGX Servers and DGX Station

MSI's AI server portfolio includes NVIDIA MGX platforms, which are tailored for scalable AI training and inference across varying thermal configurations. Additionally, the NVIDIA DGX Station, a desktop AI supercomputer, facilitates local AI development, addressing users looking for data-center-class capabilities in compact settings.

The server lineup features configurations such as the liquid-cooled 6U dual-socket AMD EPYC platform, capable of hosting up to 8 NVIDIA RTX PRO 6000 GPUs. This is complemented by the 4U dual-socket Intel Xeon platform, supporting extensive GPU setups alongside high-speed networking capabilities.

Advanced Capabilities and Infrastructure Dynamics

MSI’s infrastructure solutions focus on a balance between computational power, thermal efficiency, and flexible deployment options. Danny Hsu, General Manager of MSI's Enterprise Platform Solutions, emphasized that “scaling AI infrastructure now requires a balance between compute performance and thermal efficiency.” This philosophy drives MSI's design to support AI deployment from rack-scale infrastructures to localized AI environments.

In addition to its high-performance platforms, MSI provides multi-node hosting options under the DC-MHS architecture. This allows data centers to deploy modular infrastructure with seamless platform transitions and integration across various computing environments. The 21” Open Compute and 19” Core Compute multi-node platforms are engineered for ease of scalability and modular integration, supporting both liquid and air-cooled specifications.

Enterprise Platforms for Data Center Excellence

MSI extends its enterprise server portfolios that cover DC-MHS servers, versatile HPMs, and standard motherboard architectures, all supporting GPU-accelerated tasks for cloud and enterprise deployments. The dual-socket Intel Xeon servers and AMD EPYC systems stand out for their ability to manage expansive virtualization workloads, facilitating high-density cloud infrastructures efficiently.

As AI technology advances and becomes more integral in various sectors, MSI ensures that its infrastructure solutions are ahead of the curve, prepared to meet the challenges of tomorrow's data-intensive applications.

Conclusion

MSI's comprehensive presentation at COMPUTEX 2026 reaffirms its leadership in the AI infrastructure domain by delivering cutting-edge, liquid-cooled solutions that enable organizations to harness the full potential of their AI workloads. With a focus on performance, efficiency, and adaptability, MSI is setting the standard for the future of data centers globally.