MSI Unveils Cloud-to-Edge AI Ecosystem at COMPUTEX 2026

At its 40th-anniversary COMPUTEX showcase, MSI revealed a unified enterprise AI strategy designed to bridge the gap between massive data centers and autonomous edge devices. Rather than a collection of standalone products, the company presented its new hardware as a seamless cloud-to-edge ecosystem, with each tier built on NVIDIA's latest architectures .

Cloud Foundation: The Liquid-Cooled CG681-S6093 MGX Server

For the high end of the AI infrastructure spectrum, MSI introduced the CG681-S6093, a 6U liquid-cooled server built on the modular NVIDIA MGX architecture. It is designed to handle the immense compute density required for hyperscale AI data centers .

The server supports dual AMD EPYC processors and can be configured with up to eight NVIDIA RTX PRO 6000 Blackwell Server Edition GPUs. MSI emphasized that the liquid cooling and MGX design were specifically chosen to prioritize thermal efficiency and performance for large-scale AI training, inference, and agentic AI workloads .

Local AI Powerhouse: Manufacturing the DGX Station

Moving from the cloud to on-premises environments, MSI unveiled its role in manufacturing systems based on the NVIDIA DGX Station architecture. The XpertStation WS300 is a desktop workstation powered by the NVIDIA GB300 Grace Blackwell Ultra Desktop Superchip, supported by a massive 748 GB of unified memory .

This system is targeted at AI developers, researchers, and data scientists who need data center-class performance locally for complex AI training and fine-tuning without depending on a remote server connection .

Edge Execution: The EdgeXpert AI Supercomputer Debuts

The most novel piece of the continuum is the EdgeXpert AI Supercomputer (MS-C931). Built on the NVIDIA DGX Spark platform and driven by the NVIDIA GB10 Grace Blackwell Superchip, this device delivers 1 petaFLOP of desktop AI performance .

MSI presented the EdgeXpert as the point where data center intelligence directly meets physical operations. The company demonstrated it powering autonomous edge computing use cases, such as advanced factory inspection and real-time operational decision-making, bringing AI compute capability to places where low latency and immediate action are critical .

The EdgeXpert represents the "edge execution" pillar of the strategy, forming a complete loop with the cloud-based MGX servers for heavy lifting and DGX Station workstations for high-powered local development .