MSI, a leading global provider of high-performance server solutions, unveiled its AI server based on the NVIDIA MGX architecture at Supercomputing 2024 (SC24) from November 19-21 at booth 3655. Purpose-built to maximize compute density, energy efficiency, and modular flexibility, MSI's MGX-based AI server is designed to handle the intensive demands of AI, HPC, and data-heavy applications, offering the scalable performance and resilience that data centers need to stay ahead in today's high-performance computing landscape.

According to Danny Hsu, General Manager of MSI's Enterprise Platform Solutions, "MSI's latest innovations mark a significant leap in computational power and efficiency, enabling organizations to maximize performance, adapt seamlessly to evolving needs, and drive efficiency, building a robust foundation for future growth in high-performance computing."Leveraging NVIDIA MGX architecture, MSI's CG480-S5063 AI server accelerates deployment times and optimizes flexibility by utilizing modular building blocks that streamline engineering processes. This approach allows MSI to deliver a highly versatile AI server designed to tackle a range of compute-heavy workloads, from AI and large language models (LLMs) to complex data analytics, ensuring a future-ready infrastructure that empowers organizations to support rapidly evolving AI applications on a unified platform.

The CG480-S5063 AI server is equipped with dual Intel Xeon 6 processors and 8 FHFL dual-width GPU slots, supporting the powerful NVIDIA H200 NVL GPUs. NVIDIA H200 NVL is ideal for lower-power, air-cooled enterprise rack designs that require flexible configurations, delivering acceleration for every AI and HPC workload regardless of size. With up to four GPUs connected by NVIDIA NVLink and a 1.5X memory increase, large language model (LLM) inference can be accelerated up to 1.8X and HPC applications achieve up to 1.3X more performance over the H100 NVL.

The CG480-S5063 AI server is equipped with dual Intel Xeon 6 processors and 8 FHFL dual-width GPU slots, providing unparalleled parallel processing power for intensive AI tasks. With 32 DDR5 DIMM slots and 20 PCIe 5.0 E1.S NVMe bays, it ensures high memory bandwidth and rapid data access, while PCIe 5.0 x16 slots enable flexible and high-speed network connectivity. This 4U server delivers the performance and adaptability required for next-generation AI environments, giving organizations a cutting-edge solution for tackling even the most complex data-driven challenges.