TAIPEI, June 1, 2026 /PRNewswire/ — MSI is showcasing a full range of AI and data center infrastructure platforms at COMPUTEX 2026 (booth J0605a), built around liquid-cooled infrastructure as increasing compute density drives greater thermal requirements in modern data centers. The portfolio includes liquid-cooled ORv3 rack architectures for high-density AI infrastructure, NVIDIA MGX platforms for scalable AI training and inference, the DGX Station desktop AI supercomputer for local AI development, and DC-MHS enterprise and multi-node server platforms for Modular cloud and enterprise deployments.
Liquid-Cooled ORv3 Rack, 6U NVIDIA MGX Rack, NVIDIA DGX Station
“The evolution of AI infrastructure now requires a balance between computing performance, thermal efficiency and flexibility of deployment,” said Danny Hsu, General Manager MSI’s Enterprise Platform Solutions. « MSI’s range of AI platforms is designed to support AI deployment, from on-rack infrastructure to on-premise AI development. »
Liquid- and air-cooled rack architectures
MSI expands its infrastructure portfolio with liquid-cooled OCP ORv3 rack architectures and standard air-cooled EIA rack architectures for modern data centers. The ORv3 platform is aimed at high-density AI and cloud infrastructure, while the EIA architecture enables standardized deployment in existing enterprise environments.
Liquid-cooled 44OU ORv3 21″ rack architecture supports deployments up to 100 kW with an integrated liquid-to-liquid cooling unit (L2L CDU). Configured with 28 1OU2N Open multi-node systems Compute, the wider 21″ ORv3 design enables greater compute density, efficient liquid cooling, and 48V busbar power distribution for large-scale AI infrastructure.
The air-cooled 19″ 48RU EIA rack architecture enables standardized deployment in existing data center environments. Configured with 16 2U2N multi-node systems, the platform enables scalable cloud and enterprise infrastructure integration in standard EIA racks, with AMD EPYC™ 9005 and Intel® Xeon® 6 platform options for deployment flexibility.
Serveurs d’IA NVIDIA MGX et stations NVIDIA DGX
MSI’s NVIDIA-based AI infrastructure portfolio covers both GPU-accelerated AI servers and desktop AI supercomputer platforms. Based on the NVIDIA MGX architecture, the AI server family supports scalable AI training, inference, and HPC deployments in air- and liquid-cooled configurations. Built on the NVIDIA DGX Station architecture, the desktop platform delivers data center-grade AI computing power for local development, fine-tuning, and inference workloads.
The NVIDIA MGX server family includes 2U, 4U and 6U GPU platforms designed for AI training, inference, HPC and large data workloads. Supporting NVIDIA H200 NVL, NVIDIA RTX PRO 6000, and NVIDIA RTX PRO 4500 Blackwell Server Edition GPUs, these systems provide scalable GPU configurations in modern AI infrastructure environments, while extending MSI’s collaboration within the NVIDIA MGX ecosystem to Vera Rubin’s next-generation rack-scale platforms.
CG681-S6093: AMD EPYC 6U dual-socket liquid-cooled platform supporting up to 8 NVIDIA RTX PRO 6000 Blackwell Server Liquid Cooled Edition GPUs, 32 DDR5 DIMMs, and NVIDIA ConnectX-8 SuperNIC network interface cards with up to 8— 400 Gbps for AI training and HPC deployments.
CG480-S5063: Dual-socket 6U 4U Intel
CG481-S6053 / CG480-S6053: AMD EPYC 9005 4U dual-socket platforms supporting up to 8 double-wide GPUs, 24 DDR5 DIMMs, and 8 NVMe U.2 drives for multi-GPU AI and HPC deployments. The CG481-S6053 integrates up to 8— 400G QSFP112 via NVIDIA ConnectX-8 SuperNIC network interface cards for distributed AI training and high-bandwidth GPU clusters, with the CG480-S6053 providing 5 card slots Additional PCIe 5.0 expansion modules for scalable AI infrastructure configurations.
CG290-S3063: 2U single-socket Intel data to restricted space.
The XpertStation WS300 on the NVIDIA DGX Station architecture is a desktop AI supercomputer designed for AI development, fine-tuning, inference, and data-intensive workflows with Windows support for local development of AI applications and agents. Powered by the NVIDIA GB300 Grace Blackwell Ultra Desktop Superchip, the platform supports up to 748 GB of coherent memory and 7.1 TB/s of HBM3e bandwidth for high-speed communication between the CPU and GPU, and the processing of large AI models. Dual 400GbE networking via NVIDIA ConnectX-8 SuperNIC network interface cards and liquid-cooled thermal design enable data center-grade AI computing in a compact desktop form factor.
Plateformes multi-nœuds DC-MHS Open Compute et Core Compute
MSI offers the 21″ Open Compute and 19″ Core Compute multi-node platforms for hyperscale and cloud data centers. Built on the DC-MHS architecture, these systems enable modular platform integration, scalable infrastructure deployment, and simplified platform transition into modern IT environments.
The 21″ Open Compute family includes air- and liquid-cooled 1OU2N, 2OU2N, and 2OU4N platforms optimized for high-density AI and cloud infrastructure with 48V DC bus power distribution.
CD281-S4051-X4 (Liquid Cooled): A 2 OR 4 node liquid cooled platform supporting a single AMD EPYC 9005 CPU, 12 DDR5 DIMMs and 4 NVMe E1.S bays per node for AI inference and cloud-native infrastructure.
CD281-S4051-X2: 2OR 2-node platform supporting a single AMD EPYC 9005 processor, 12 DDR5 DIMMs, 12 E3.S NVMe bays, and two PCIe 5.0 FHHL expansion card slots per node for storage-intensive cloud and scale-out deployments.
The 19″ Core Compute family includes 2U2N and 2U4N platforms for deploying enterprise and cloud infrastructure in standard 19″ EIA rack environments.
CD270-S3071-X4 / CD270-S3071-X2: 2U 4-node and 2U 2-node platforms based on Intel Xeon 6/6+ with 12 DDR5 DIMMs per node, supporting Intel Xeon 6+ processors with up to 288 E-cores. The X4 supports 3 NVMe U.2 bays per node for compute-oriented cloud infrastructure, with the X2 supporting 6 bays per node for virtualization and data-oriented services.
CD270-S3061-X4: Intel Xeon 6 2U 4-node platform supporting 16 DDR5 DIMMs and 3 NVMe U.2 bays per node for scale-out and containerized deployments.
CD270-S4051-X4 / CD270-S4051-X2: 2U 4-node and 2U 2-node platforms based on AMD EPYC 9005 supporting 12 DDR5 DIMMs per node. X4 supports 3 NVMe U.2 bays per node for infrastructure at high computing density, the X2 adding 6 NVMe U.2 bays per node for mixed computing and storage environments.
Enterprise platforms for DC-MHS and standard architectures
MSI’s enterprise platform portfolio includes DC-MHS enterprise servers, modular host processor modules, and standard enterprise motherboards for deploying cloud and enterprise infrastructure. The portfolio includes GPU server platforms, modular DC-MHS platform integration, and standard server motherboards designed for flexible deployment in enterprise and workstation environments.
DC-MHS Enterprise Servers
CX270-S5062 (-HE SKU) / CX170-S5062: 2U and 1U dual-socket platforms based on Intel Xeon 6 and supporting 32 DDR5 DIMMs. The CX270-S5062 supports 8 NVMe U.2 drives and up to 2 dual-width 600W GPUs for virtualization and GPU-accelerated workloads, with the CX170-S5062 supporting 12 NVMe U.2 drives for high-density cloud infrastructure.
CX271-S4056 (-HE SKU) / CX171-S4056: 2U and 1U single-socket platforms based on AMD EPYC 9005 and supporting 24 DDR5 DIMMs. The CX271-S4056 supports 8 NVMe U.2 drives and up to 2 double-wide 600W GPUs for AI inference and accelerated computing, with the CX171-S4056 supporting 12 NVMe U.2 drives for scale-out cloud infrastructure.
CX271-S3066 (-HE SKU) / CX171-S3066: 2U and 1U single-socket platforms based on Intel Xeon 6 and supporting 16 DDR5 DIMMs. The CX271-S3066 supports 8 NVMe U.2 drives and up to 2 dual-width 600W GPUs, with the CX171-S3066 supporting 12 NVMe U.2 drives for consumer cloud deployments.
DC-MHS host processor module and standard enterprise motherboards
Supporting the M-DNO Type-2, M-DNO Type-4 and M-FLW form factors, the DC-MHS HPM family provides modular integration of the DC-MHS platform, with standard server motherboards supporting popular enterprise and workstation platforms.
View original content:https://www.prnewswire.com/news-releases/msi-presente-linfrastructure-ia-a-refroidissement-liquide-les-plateformes-nvidia-mgx-nvidia-dgx-station-et-dc-mhs-au-computex-2026-302786327.html
