- Joined
- May 21, 2024
- Messages
- 1,341 (3.66/day)
MSI, a leading global provider of high-performance server solutions, unveils its latest AI innovations using NVIDIA MGX and NVIDIA DGX Station reference architectures at COMPUTEX 2025, held from May 20-23 at booth J0506. Purpose-built to address the growing demands of AI, HPC, and accelerated computing workloads, MSI's AI solutions feature modular, scalable building blocks designed to deliver next-level AI performance for enterprises and cloud data center environments.
"AI adoption is transforming enterprise data centers as organizations move quickly to integrate advanced AI capabilities," said Danny Hsu, General Manager of Enterprise Platform Solutions at MSI. "With the explosive growth of generative AI and increasingly diverse workloads, traditional servers can no longer keep pace. MSI's AI solutions, built on the NVIDIA MGX and NVIDIA DGX Station reference architectures, deliver the scalability, flexibility, and performance enterprises need to future-proof their infrastructure and accelerate their AI innovation."
MSI showcases a rack solution built on the NVIDIA Enterprise Reference Architecture, featuring a four-node scalable unit—an MSI AI server built on NVIDIA MGX. Each server is equipped with eight NVIDIA H200 NVL GPUs and enhanced by the NVIDIA Spectrum -X networking platform to provide leading scalability for AI workloads. This modular architecture can scale up to 32 server systems (eight scalable units), supporting a total of 256 NVIDIA H200 NVL GPUs. MSI's Enterprise Reference Architecture is optimized for multi-node AI and hybrid applications, designed to tackle today's most demanding computational challenges and drive the next generation of data center innovation.
Leveraging the NVIDIA MGX modular architecture, MSI's AI servers provide a foundation for accelerated computing and meet the diverse demands of AI, HPC, and NVIDIA Omniverse workloads. MSI's 4U AI server offers configuration options with both Intel and AMD CPUs and is a cutting-edge solution designed for large language models (LLMs), deep learning training, and fine-tuning. The CG480-S5063 AI server platform features dual Intel Xeon 6 processors and eight FHFL dual-width GPU slots, supporting both NVIDIA H200 NVL and NVIDIA RTX PRO 6000 Blackwell Server Edition with power capacities of up to 600 W. Featuring 32 DDR5 DIMM slots and twenty PCIe 5.0 E1.S NVMe bays, the CG480-S5063 ensures exceptional memory bandwidth and rapid data access. Its modular design and robust storage capabilities make it the perfect choice for next-generation AI and HPC applications, delivering unparalleled performance and scalability.
The CG290-S3063 is a 2U AI server platform built on the NVIDIA MGX architecture. It features a single-socket Intel Xeon 6 processor, 16 DDR5 DIMM slots, and four FHFL dual-width GPU slots supporting power capacities of up to 600 W. With PCIe 5.0 expansion slots, four rear 2.5-inch NVMe drive bays, and two M.2 NVMe slots, this server delivers an ideal solution for a wide range of AI workloads—from small-scale inferencing to large-scale AI training.
MSI's platforms support deployment in enterprise-ready AI environments with the NVIDIA Enterprise AI Factory validated design that provides guidance for developing, deploying, and managing agentic AI, physical AI, and HPC workloads on the NVIDIA Blackwell platform with on-premises infrastructure. Designed for enterprise IT, the validated design includes accelerated computing, networking, storage, and software to help deliver faster time-to-value for AI factory deployments while mitigating deployment risks.
Also on display is MSI's AI Station CT60-S8060, featuring the powerful components of the NVIDIA DGX Station, designed to deliver data center-level performance on a personal desktop for AI development, featuring the NVIDIA GB300 Grace Blackwell Ultra Desktop Superchip, and up to 784 GB of coherent memory to accelerate large-scale training and inferencing workloads. Combining state-of-the-art system capabilities with the NVIDIA AI Enterprise software stack, DGX Station is purpose-built for teams that demand the best desktop AI development solution.
View at TechPowerUp Main Site | Source
"AI adoption is transforming enterprise data centers as organizations move quickly to integrate advanced AI capabilities," said Danny Hsu, General Manager of Enterprise Platform Solutions at MSI. "With the explosive growth of generative AI and increasingly diverse workloads, traditional servers can no longer keep pace. MSI's AI solutions, built on the NVIDIA MGX and NVIDIA DGX Station reference architectures, deliver the scalability, flexibility, and performance enterprises need to future-proof their infrastructure and accelerate their AI innovation."
MSI showcases a rack solution built on the NVIDIA Enterprise Reference Architecture, featuring a four-node scalable unit—an MSI AI server built on NVIDIA MGX. Each server is equipped with eight NVIDIA H200 NVL GPUs and enhanced by the NVIDIA Spectrum -X networking platform to provide leading scalability for AI workloads. This modular architecture can scale up to 32 server systems (eight scalable units), supporting a total of 256 NVIDIA H200 NVL GPUs. MSI's Enterprise Reference Architecture is optimized for multi-node AI and hybrid applications, designed to tackle today's most demanding computational challenges and drive the next generation of data center innovation.
Leveraging the NVIDIA MGX modular architecture, MSI's AI servers provide a foundation for accelerated computing and meet the diverse demands of AI, HPC, and NVIDIA Omniverse workloads. MSI's 4U AI server offers configuration options with both Intel and AMD CPUs and is a cutting-edge solution designed for large language models (LLMs), deep learning training, and fine-tuning. The CG480-S5063 AI server platform features dual Intel Xeon 6 processors and eight FHFL dual-width GPU slots, supporting both NVIDIA H200 NVL and NVIDIA RTX PRO 6000 Blackwell Server Edition with power capacities of up to 600 W. Featuring 32 DDR5 DIMM slots and twenty PCIe 5.0 E1.S NVMe bays, the CG480-S5063 ensures exceptional memory bandwidth and rapid data access. Its modular design and robust storage capabilities make it the perfect choice for next-generation AI and HPC applications, delivering unparalleled performance and scalability.
The CG290-S3063 is a 2U AI server platform built on the NVIDIA MGX architecture. It features a single-socket Intel Xeon 6 processor, 16 DDR5 DIMM slots, and four FHFL dual-width GPU slots supporting power capacities of up to 600 W. With PCIe 5.0 expansion slots, four rear 2.5-inch NVMe drive bays, and two M.2 NVMe slots, this server delivers an ideal solution for a wide range of AI workloads—from small-scale inferencing to large-scale AI training.
MSI's platforms support deployment in enterprise-ready AI environments with the NVIDIA Enterprise AI Factory validated design that provides guidance for developing, deploying, and managing agentic AI, physical AI, and HPC workloads on the NVIDIA Blackwell platform with on-premises infrastructure. Designed for enterprise IT, the validated design includes accelerated computing, networking, storage, and software to help deliver faster time-to-value for AI factory deployments while mitigating deployment risks.
Also on display is MSI's AI Station CT60-S8060, featuring the powerful components of the NVIDIA DGX Station, designed to deliver data center-level performance on a personal desktop for AI development, featuring the NVIDIA GB300 Grace Blackwell Ultra Desktop Superchip, and up to 784 GB of coherent memory to accelerate large-scale training and inferencing workloads. Combining state-of-the-art system capabilities with the NVIDIA AI Enterprise software stack, DGX Station is purpose-built for teams that demand the best desktop AI development solution.
View at TechPowerUp Main Site | Source