MSI EdgeXpert MS-C931 AI Supercomputer – NVIDIA GB10 Grace Blackwell, 128GB Unified Memory, 1 PetaFLOP AI Performance
Key Highlights
- NVIDIA GB10 Grace Blackwell Superchip with Blackwell GPU architecture
- 20-core Arm CPU with 10 Cortex-X925 and 10 Cortex-A725 cores
- 128GB LPDDR5x coherent unified system memory
- Up to 1 petaFLOP of FP4 sparse AI computing performance
The MSI EdgeXpert MS-C931 is a compact desktop AI supercomputer designed for AI developers, researchers, data scientists and enterprise applications. Powered by the NVIDIA GB10 Grace Blackwell Superchip, it delivers up to 1 petaFLOP of AI computing performance in a compact desktop form factor.
Combining a 20-core Arm CPU with NVIDIA Blackwell GPU architecture, the EdgeXpert features 128GB of LPDDR5x coherent unified system memory, allowing users to develop, fine-tune and run advanced AI models locally.
Built on the NVIDIA DGX Spark platform, it comes with NVIDIA DGX OS and a preconfigured NVIDIA AI software stack, including support for popular development tools such as PyTorch, Jupyter and Ollama.
With NVIDIA ConnectX-7 networking supporting up to 200GbE, Wi-Fi 7, Bluetooth and high-speed USB-C connectivity, the EdgeXpert is designed for demanding AI workloads, local inference and scalable multi-system deployments.
BOX CONTENTS
- MSI EdgeXpert MS-C931 AI Supercomputer
- External power adapter
- Power cable
- User documentation
WARRANTY
Warranty: 1 year
FAQ
What is the MSI EdgeXpert MS-C931?
It is a compact desktop AI supercomputer built on the NVIDIA DGX Spark platform and powered by the NVIDIA GB10 Grace Blackwell Superchip.
Which processor does it use?
It features a 20-core Arm CPU consisting of 10 Cortex-X925 and 10 Cortex-A725 cores.
Which GPU architecture does it use?
It uses NVIDIA Blackwell GPU architecture integrated into the GB10 Grace Blackwell Superchip.
How much memory does it have?
It features 128GB of LPDDR5x coherent unified system memory shared between the CPU and GPU.
What is its AI computing performance?
It delivers up to 1 petaFLOP of AI performance using FP4 sparse tensor computing.
What operating system does it run?
It runs NVIDIA DGX OS, an Ubuntu Linux-based operating system with the NVIDIA AI software stack preinstalled.
Can it run AI models locally?
Yes. The 128GB unified memory configuration supports running AI models with up to approximately 200 billion parameters locally, depending on model requirements and configuration.
Can multiple EdgeXpert systems work together?
Yes. Multiple systems can be connected using NVIDIA ConnectX networking to scale AI computing and memory capacity for larger models.
What storage does it support?
It supports self-encrypting NVMe M.2 SSD storage. Available configurations include 1TB and 4TB, depending on the supplied model.
What connectivity options are available?
It includes four USB 3.2 Type-C ports, HDMI 2.1a, 10GbE Ethernet, NVIDIA ConnectX-7 networking, Wi-Fi 7 and Bluetooth.
Is it suitable for gaming?
The EdgeXpert is primarily designed for AI development, inference, research and professional computing rather than conventional gaming.
| Brand | MSI |
| Model Number | MS-C931 |
| Product Type | Desktop AI Supercomputer |
| Platform | NVIDIA DGX Spark |
| Architecture | NVIDIA Grace Blackwell |
| Processor | NVIDIA GB10 Grace CPU |
| CPU Cores | 20-Core Arm CPU |
| CPU Configuration | 10 × Cortex-X925 + 10 × Cortex-A725 |
| GPU Architecture | NVIDIA Blackwell |
| AI Performance | Up to 1 PetaFLOP (FP4 Sparse) |
| System Memory | 128GB |
| Memory Type | LPDDR5x Coherent Unified Memory |
| Memory Interface | 256-bit |
| Memory Bandwidth | 273 GB/s |
| Maximum AI Model Size | Up to 200 Billion Parameters |
| Storage Type | NVMe M.2 SSD |
| Storage Options | 1TB / 4TB (Configuration Dependent) |
| Operating System | NVIDIA DGX OS |
| Networking | NVIDIA ConnectX-7 |
| Maximum Network Speed | Up to 200GbE |
| Ethernet | 1 × RJ-45 10GbE |
| USB Ports | 4 × USB 3.2 Type-C |
| HDMI | 1 × HDMI 2.1a |
| Display Support | Up to 4 Independent Displays |
| DisplayPort | 3 × DisplayPort via USB-C |
| Wireless Connectivity | Wi-Fi 7 |
| Bluetooth | Bluetooth 5.4 |
| Power Adapter | 240W USB-C |
| Dimensions | 151 × 151 × 52 mm |
| Volume | 1.19 Litres |
| Weight | 1.2 kg |
| Color | Black |
| Usage | AI Development, Machine Learning, Deep Learning, AI Inference and Research |
| Condition | New |
