- Brand: MSI
- Model: EdgeXpert 4TB Dual Pack
- EAN: 4711377407434
- MPN: 9S6-C9311-65S
- NVIDIA Grace Blackwell Architecture
- NVIDIA Blackwell GPU & Arm 20-core CPU
- 20 core Arm, 10 Cortex-X925 + 10 Cortex-A725 Arm
- NVIDIA® NVLink®-C2C CPU-GPU memory interconnect
- 128 GB LPDDR5x coherent, unified system memory
- 1000 AI FLOPS (FP4) AI performance
- Full stack solution, hardware & software, designed for AI developers
- MSI EdgeXpert GB10 4TB x 2 Units
- 3 Years Warranty
- Dispatch after 2 days
Ask a Question About This Product
MSI EdgeXpert 4TB Dual Pack
Personal AI Supercomputer Based on NVIDIA Grace Blackwell GB10
The MSI EdgeXpert 4TB Dual Pack is a two-system desktop AI supercomputer solution designed for AI developers, researchers, data scientists, academic labs, enterprise AI teams and edge AI deployments. Based on the NVIDIA Grace Blackwell GB10 platform, each EdgeXpert system delivers 1000 AI TOPS FP4 performance, 128GB LPDDR5x unified system memory, 4TB NVMe M.2 self-encrypting storage, NVIDIA DGX OS and high-speed ConnectX-7 networking for local AI development, inference, fine-tuning and scalable multi-system workflows.
Key Highlights
A compact dual-system AI platform built for local LLM development, private AI inference, edge AI applications and scalable AI research workflows.
Grace Blackwell
NVIDIA Grace Blackwell GB10 architecture with Blackwell GPU and Arm 20-core CPU.
AI TOPS
1000 AI TOPS FP4 tensor performance per EdgeXpert system.
Unified Memory
128GB LPDDR5x coherent unified system memory for large AI workflows.
NVMe Storage
4TB NVMe M.2 self-encrypting SSD storage per system in this configuration.
Dual Scaling
Two connected EdgeXpert systems can work with AI models up to 405B parameters.
ConnectX-7
ConnectX-7 Smart NIC with up to 200Gbps high-speed system networking.
DGX OS
NVIDIA DGX OS platform for AI development and deployment workflows.
Compact Design
151 x 151 x 52 mm compact 1.19L system design for desktop AI setups.
Next-Level AI Power, Right at Your Desk
MSI EdgeXpert brings petaflop-class AI performance into a compact desktop-ready form factor. It is designed for users who need local AI compute for building, testing, fine-tuning and running AI models without relying completely on public cloud infrastructure.
Compact AI Supercomputer for Local AI Development
The EdgeXpert MS-C931 is built for desktop AI development, private AI inference, regulated environments, academic labs and edge AI applications where local compute, data control and fast experimentation matter.
Dual Pack Advantage for Larger AI Workflows
This Dual Pack includes two MSI EdgeXpert 4TB systems. Each system provides independent GB10 compute, 128GB unified memory and 4TB NVMe storage. Using high-speed NVIDIA ConnectX networking, two EdgeXpert systems can work together for larger AI models and advanced local AI workflows up to 405 billion parameters.
Two-System AI Development Setup
A dual-pack configuration gives AI teams more flexibility for parallel experimentation, multi-user lab setups, distributed testing, large model evaluation and advanced local AI development workflows.
405B Parameter Scaling
MSI highlights that two EdgeXpert systems connected through NVIDIA ConnectX networking can work with AI models up to 405 billion parameters, making it suitable for high-end local LLM and research environments.
NVIDIA Grace Blackwell GB10 Architecture
At the core of EdgeXpert is the NVIDIA Grace Blackwell GB10 platform, combining NVIDIA Blackwell GPU architecture with an Arm 20-core CPU. The NVLink-C2C CPU-GPU memory interconnect helps provide a unified memory model for efficient AI development, data preprocessing, orchestration and real-time inference workflows.
Blackwell GPU + Arm 20-Core CPU
The GB10 platform combines NVIDIA Blackwell GPU acceleration with a 20-core Arm CPU, helping developers handle AI model tuning, inference and data workflows efficiently on a compact desktop system.
NVLink-C2C Memory Interconnect
NVLink-C2C creates a seamless CPU and GPU memory model, helping improve data movement efficiency for AI model development, inference and high-performance local AI workloads.
Large Unified Memory for Local LLMs
Each EdgeXpert system includes 128GB LPDDR5x unified system memory with 256-bit memory interface and 273GB/s memory bandwidth. This large memory architecture helps developers run AI models locally, reduce cloud dependency, improve data privacy and maintain better control over AI experimentation.
128GB LPDDR5x
Large unified memory for AI development, testing and inference workloads.
273GB/s
High memory bandwidth for demanding AI and data processing tasks.
Single System
Run AI models up to 200B parameters locally on one EdgeXpert system.
Dual System
Scale to AI models up to 405B parameters with two connected systems.
High-Speed Connectivity for AI Workstations
EdgeXpert is equipped for professional AI development environments with USB Type-C connectivity, HDMI output, 10GbE networking, Wi-Fi 7, Bluetooth 5.4 and ConnectX-7 Smart NIC support. The rear I/O is designed for modern desktop AI labs, multi-display workflows and high-speed data movement.
Professional Rear I/O
EdgeXpert includes 4 x USB 3.2 Type-C, HDMI 2.1a, 1 x 10GbE RJ-45, QSFP ports for ConnectX networking, Wi-Fi antennas and external power input for compact AI workstation deployment.
ConnectX-7 Smart NIC
The integrated ConnectX-7 Smart NIC supports high-speed networking up to 200Gbps, helping dual EdgeXpert systems scale for larger local AI model workloads and advanced development scenarios.
Scale AI Models from Desktop to Cloud
MSI positions EdgeXpert as a compact AI platform that helps developers build locally and then scale workflows to NVIDIA DGX Cloud, DGX infrastructure or other NVIDIA-accelerated data centers with minimal code changes. This makes it suitable for teams that want local control during development with future cloud or data center migration.
Local AI Development with Cloud-Ready Scaling
Develop, validate and optimize AI applications locally, then migrate AI programming to NVIDIA DGX Cloud or NVIDIA-accelerated infrastructure when larger compute resources are needed.
Designed for AI Developers, Education and Independent AI Builders
EdgeXpert is designed for professionals and institutions working on AI model development, local inference, academic AI training and independent AI projects. It supports AI workflows from prototyping and fine-tuning to private local inference and application testing.
Enterprise AI Developers
Suitable for data scientists and AI developers who need to fine-tune models, build inference applications locally and migrate workloads to company data centers or cloud infrastructure.
Independent AI Development
Ideal for advanced independent developers experimenting with local inference, AI assistants, model testing and AI application creation in a controlled desktop environment.
Supported AI Framework Applications
EdgeXpert provides a strong platform for edge AI applications powered by NVIDIA AI frameworks, including use cases across medical, robotics, retail and smart city workloads.
Medical AI
Supports NVIDIA Holoscan-based medical AI and device development workflows.
Robotics AI
Suitable for NVIDIA Isaac-based robot and physical AI development.
Retail AI
Useful for retail AI workloads including voice, semantic and edge inference.
Smart City
Supports NVIDIA Metropolis-based smart city and video analytics projects.
Robotics and Automation AI
EdgeXpert can support robotics, automation, simulation and real-time AI decision-making workflows where local AI acceleration and edge deployment are important.
Smart City and Vision AI
Ideal for smart city, camera analytics, traffic intelligence and visual AI development using NVIDIA accelerated AI frameworks.
Enterprise On-Prem AI Advantage
EdgeXpert is not only a compact AI supercomputer; it is also positioned for secure, private and enterprise-ready AI deployment. MSI highlights on-premise AI use cases such as legal AI, machine vision, offline voice AI and regulated industry AI workloads where data control and local processing matter.
On-Prem AI
Run sensitive AI workloads locally without relying fully on public cloud services.
Legal AI
Supports contract review, document analysis and internal knowledge reasoning use cases.
Machine Vision
Useful for factory inspection, QA, SOP guidance and visual AI workflows.
Offline Voice AI
Supports secure local voice AI translation and semantic AI deployments.
Specification Overview
| Product Name | MSI EdgeXpert 4TB Dual Pack |
|---|---|
| Bundle Type | 2 x MSI EdgeXpert MS-C931 4TB Systems |
| Part Number | EdgeXpert MS-C931 |
| Architecture | NVIDIA Grace Blackwell GB10 |
| GPU | NVIDIA Blackwell Architecture |
| CPU | 20-core Arm CPU, 10 Cortex-X925 + 10 Cortex-A725 |
| CPU Type | Onboard |
| Tensor Performance | 1 petaFLOP AI Performance / 1000 AI TOPS FP4 Sparse per system |
| Memory Technology | LPDDR5x |
| Maximum Memory | 128GB unified system memory per system |
| Graphics Controller | NVIDIA GB10 |
| Graphics Memory | Shared 128GB unified memory |
| Single-System Model Support | AI models up to 200 billion parameters |
| Dual-System Model Support | AI models up to 405 billion parameters with two connected systems |
| Storage | 4TB NVMe M.2 self-encrypting SSD per system |
| Total Bundle Storage | 2 x 4TB systems, storage works per individual system |
| USB | 4 x USB 3.2 Type-C |
| Ethernet | 1 x RJ-45 10GbE |
| NIC | ConnectX-7 Smart NIC @ 200Gbps |
| Wi-Fi | Wi-Fi 7 |
| Bluetooth | Bluetooth 5.4 |
| Audio Output | HDMI multichannel audio output |
| Display Connectors | 1 x HDMI 2.1a, DisplayPort via USB-C |
| Multiple Display | 4 independent displays |
| HDMI | 1 |
| DisplayPort | 3 via USB-C |
| QSFP | 2 |
| NVENC / NVDEC | 1 x NVENC / 1 x NVDEC |
| Power Input | ~240W USB-C with external power adapter per system |
| Smart Fan | Yes |
| TPM | Yes |
| Operating Temperature | 0°C to 35°C |
| Operating Humidity | 10% to 90%, non-condensing |
| Storage Temperature | 0°C to 60°C |
| Dimensions | 151 x 151 x 52 mm per system |
| Volume | 1.19L per system |
| Weight | 1.2 kg per system |
| Certifications | FCC Class B, CE, UKCA, VCCI, BSMI |
| Operating System | NVIDIA DGX OS |
| Best For | AI development, local LLM inference, fine-tuning, data science, machine vision, robotics, smart city AI and on-prem enterprise AI |
FAQ
What is MSI EdgeXpert 4TB Dual Pack?
It is a two-system bundle of MSI EdgeXpert MS-C931 AI supercomputers, each configured with 4TB NVMe M.2 self-encrypting storage for local AI development and edge AI workloads.
Which AI platform powers EdgeXpert?
MSI EdgeXpert is based on the NVIDIA Grace Blackwell GB10 platform with NVIDIA Blackwell GPU architecture and a 20-core Arm CPU.
How much memory does each system include?
Each EdgeXpert system includes 128GB LPDDR5x unified system memory.
How much AI performance does it deliver?
MSI lists the system with 1 petaFLOP AI performance / 1000 AI TOPS FP4 Sparse tensor performance per system.
Can two EdgeXpert systems work together?
Yes. MSI states that two connected EdgeXpert systems can work with AI models up to 405 billion parameters using NVIDIA ConnectX networking.
What type of AI workloads is this product suitable for?
It is suitable for local LLM inference, fine-tuning, data science, machine vision, robotics AI, medical AI, smart city AI, legal AI and secure on-prem enterprise AI workloads.
Run Private AI, Local LLMs and Edge AI Workloads with Dual EdgeXpert Power
Choose the MSI EdgeXpert 4TB Dual Pack for a compact, scalable and enterprise-ready desktop AI supercomputer solution built for local AI development, private inference, model fine-tuning, edge AI and regulated on-premise AI deployments.
