MSI EdgeXpert MS-C931 – AI Supercomputer Based on NVIDIA DGX Spark Platform for Deep Learning
4 TB
$6,499.00
Skip to product information
MSI EdgeXpert MS-C931 – AI Supercomputer Based on NVIDIA DGX Spark Platform for Deep Learning
$6,499.00
The MSI EdgeXpert MS-C931 is a desktop-class AI supercomputer built on the NVIDIA DGX Spark platform. It suits developers, researchers and data scientists who want data-center-grade AI capability on their own desk instead of metered by the hour in the cloud. It pairs the NVIDIA GB10 Grace Blackwell Superchip with 128GB of unified memory in a 1.19-litre chassis, and MSI adds a curated software ecosystem.
Who it's for
- Developers and researchers fine-tuning, quantizing and serving models up to roughly 200B parameters
- Teams building legal AI, machine vision, offline voice translation or agentic AI products with the bundled software
- Teams with data-residency or IP concerns keeping datasets and model weights on-premises
Key features
- NVIDIA GB10 Grace Blackwell Superchip: a 20-core Arm CPU (10x Cortex-X925 + 10x Cortex-A725) fused to a Blackwell-architecture GPU over NVIDIA NVLink-C2C, delivering 1000 AI TOPS / 1 petaFLOP of FP4 (sparse) performance.
- 128GB unified LPDDR5x memory at 273GB/s: a single coherent 256-bit memory pool shared between CPU and GPU, holding models up to roughly 200B parameters (and up to 405B across two stacked units via ConnectX-7).
- 1TB or 4TB self-encrypting NVMe SSD: hardware self-encryption at either tier.
- NVIDIA ConnectX-7 Smart NIC (200Gbps): data-center-class networking for clustering two units, plus 10GbE, Wi-Fi 7 and Bluetooth 5.4.
- NVIDIA DGX OS pre-installed: the full NVIDIA AI stack ships ready to run, and work is container-portable to other DGX Spark-class systems or DGX Cloud.
- Curated third-party AI software: MSI ships access to AIPLUX Legal AI Suite, Memorence Operagents (machine vision and operational AI), Ubestream offline voice translation and Galene.AI Elettra’s agentic AI platform.
- Compact and quiet: a 151 x 151 x 52mm, 1.2kg chassis with smart fan control, 4x USB 3.2 Type-C, HDMI 2.1a and 3x DisplayPort 1.4a.
- Enterprise reliability: TPM onboard, a 0–35°C operating temperature range, and FCC/CE/UKCA/VCCI/BSMI certification.
Specifications
| Spec | Detail |
|---|---|
| Model Number | EdgeXpert MS-C931 |
| Architecture | NVIDIA Grace Blackwell GB10 Superchip |
| CPU | 20-core Arm (10x Cortex-X925 + 10x Cortex-A725) |
| CPU-GPU Interconnect | NVIDIA NVLink-C2C |
| Tensor Performance | 1 petaFLOP (1000 AI TOPS), FP4, Sparse |
| System Memory | 128GB LPDDR5x coherent unified system memory |
| Memory Interface / Bandwidth | 256-bit / 273GB/s |
| Storage | 1TB or 4TB NVMe M.2 SSD with self-encryption |
| USB | 4x USB 3.2 Type-C |
| Ethernet | 1x RJ-45 (10GbE) |
| Network Interface Card | NVIDIA ConnectX-7 Smart NIC (200Gbps) |
| Display Connectors | 1x HDMI 2.1a; 3x DisplayPort 1.4a via USB-C |
| Video Encode/Decode | 1x NVENC / 1x NVDEC |
| TPM / Smart Fan Control | Yes / Yes |
| Operating System | NVIDIA DGX OS |
| Max Model Size | Up to 200B parameters (405B via dual-unit ConnectX stacking) |
| Dimensions | 151 x 151 x 52mm (1.19L) |
| Weight | 1.2kg |
| Operating Temperature / Humidity | 0–35°C / 10–90% non-condensing |
| Storage Temperature / Humidity | 0–60°C / 10–90% non-condensing |
| Power Supply | ~240W via USB-C with external adapter |
| Certifications | FCC Class B, CE, UKCA, VCCI, BSMI |
| Bundled AI Software | AIPLUX Legal AI Suite, Memorence Operagents, Ubestream Voice Translation, Galene.AI Elettra |
| In the Box | EdgeXpert MS-C931 system, external ~240W USB-C power adapter, power cord, quick start guide |
Good to know
- The 1 petaFLOP / 1000 TOPS figure is a peak FP4-sparse number. Real dense-precision throughput on a given model will be lower.
- Training a foundation model from scratch is still cluster or cloud territory. This machine is built for fine-tuning, quantization, inference serving and applied-AI development.
- A 70B model in FP4/INT4 quantization, or a 30B model at higher precision with a long context window, runs without memory workarounds.
- Choose 1TB for a handful of active checkpoints and datasets, or 4TB for multiple quantized variants, larger training corpora or a growing library of fine-tuned checkpoints.
Pick the MS-C931 over a bare DGX Spark unit if the bundled software overlaps with what you are building.