MSI EdgeXpert-32SEU – Personal AI Supercomputer Powered by NVIDIA Grace Blackwell

Bring data center-class AI capabilities directly to your desktop with the MSI EdgeXpert-32SEU. Built around the powerful NVIDIA GB10 Grace Blackwell Superchip, EdgeXpert is a compact AI development platform designed for developers, researchers, data scientists and professionals working with advanced AI workloads.

With up to 1 petaFLOP of FP4 AI performance, 128 GB LPDDR5x coherent unified memory and a high-speed 4 TB NVMe SSD, EdgeXpert provides the resources required to develop, fine-tune and run sophisticated AI models locally.

NVIDIA Grace Blackwell AI Performance

At the heart of EdgeXpert is the NVIDIA GB10 Grace Blackwell Superchip, combining NVIDIA Blackwell GPU technology with a powerful 20-core Arm CPU.

The integrated NVIDIA NVLink-C2C interconnect creates a coherent CPU and GPU memory architecture, enabling fast access to the system's large unified memory pool and providing exceptional performance for modern AI workloads.

Key Highlights

• NVIDIA GB10 Grace Blackwell Superchip
• Up to 1 petaFLOP FP4 AI performance
• 20-core Arm CPU
• 128 GB LPDDR5x coherent unified memory
• 4 TB high-speed NVMe SSD
• NVIDIA DGX OS based on Ubuntu Linux
• NVIDIA AI software stack pre-installed
• ConnectX-7 high-speed networking
• Up to 200 GbE connectivity
• Wi-Fi 7 and Bluetooth 5.3
• Compact desktop form factor

Run Large AI Models Locally

The massive 128 GB unified memory architecture enables EdgeXpert to handle AI workloads that are difficult to run on conventional desktop PCs and workstations.

A single EdgeXpert system can run AI models with up to 200 billion parameters locally, making it highly suitable for large language models, AI agents, multimodal AI, inference, fine-tuning and data science applications.

Local AI processing also enables developers to keep models and datasets close to their development environment rather than relying entirely on cloud-based computing resources.

Complete NVIDIA AI Development Platform

EdgeXpert comes with NVIDIA DGX OS, based on Ubuntu Linux, together with NVIDIA's AI software stack.

Popular development environments and frameworks including PyTorch, Jupyter and Ollama are supported, giving AI developers a powerful platform for prototyping, development, fine-tuning and local inference.

Scale Beyond a Single System

EdgeXpert is designed to grow with more demanding AI workloads.

With integrated NVIDIA ConnectX-7 networking, multiple EdgeXpert systems can be interconnected to create powerful desktop AI clusters.

Up to four systems can be combined, providing as much as:

• 512 GB combined unified memory
• 4 petaFLOPs FP4 AI compute
• Support for AI models up to 700 billion parameters

This provides a flexible path from individual AI development to significantly larger multi-node workloads.

Compact AI Power for Professional Environments

Despite its exceptional AI capabilities, EdgeXpert uses an extremely compact desktop design. Advanced thermal engineering, including MSI's FanDock airflow technology and dedicated SSD Vapor Chamber cooling, is designed to maintain performance during sustained AI workloads.

The result is a powerful local AI development platform that brings the NVIDIA Grace Blackwell architecture directly to the desktop.

Ideal For

• AI and machine learning development
• Large Language Models (LLMs)
• Generative AI
• AI agents
• Multimodal AI
• Local AI inference
• Model fine-tuning
• Data science
• Research and development
• Edge AI development
• AI prototyping

MSI EdgeXpert-32SEU combines NVIDIA Grace Blackwell computing, massive unified memory and the NVIDIA AI software ecosystem in a compact desktop system built specifically for the next generation of AI development.