MSI has launched a new range of AI and high-performance computing (HPC) server solutions designed to enhance computational power, efficiency, and scalability. These servers feature the NVIDIA MGX architecture and Intel Xeon 6 processors, addressing the demanding requirements of AI, HPC, and data-intensive applications.

The new AI servers, such as the MSI CG480-S5063, integrate dual Intel Xeon 6 processors with eight GPU slots and support NVIDIA H200 NVL GPUs, making them ideal for AI workloads, large language models, and data analytics. These systems are equipped with 32 DDR5 DIMM slots and PCIe 5.0 slots to ensure high throughput and flexibility, vital for AI and data-centric environments.

MSI’s DC-MHS server solutions are crafted to offer scalable, modular designs with high-density computing to meet evolving needs in data centers. These systems are powered by Intel Xeon 6 processors, combining performance-boosting P-cores with energy-efficient E-cores for optimized resource management. Enhanced cooling features, such as the Extended Volume Air Cooling (EVAC) system, enable stable operations under heavy workloads.

Key models include the 2U servers like the CD270-S3061-X2, which is ideal for memory-intensive applications, and the CX270-S5062, designed for environments requiring high compute throughput with dual GPUs. These solutions are equipped with ample PCIe slots and support for NVMe storage, which is crucial for high-speed data processing.

MSI also offers server motherboards with enhanced scalability, such as the D3071 M-DNO and D5062 M-FLW, which provide support for multiple Xeon processors and DDR5 memory configurations. These components are designed for flexibility, supporting HPC, AI, and data analytics needs across industries. Additionally, the MGT1 DC-SCM2 Module enables cross-platform compatibility, which reduces deployment and maintenance costs.

Overall, MSI’s new server offerings are poised to advance AI, HPC, and data analytics, providing organizations with scalable, efficient solutions for demanding computational tasks.