Exxact Valence DGX Station
Use this to tell your different configurations apart in your cart — handy if you order the same product with different options.
Full Specifications
Compute & Memory
Superchip
NVIDIA GB300 Grace Blackwell Ultra Desktop Superchip
CPU
1× NVIDIA Grace CPU Superchip, 72 Arm Neoverse V2 cores
GPU
1× NVIDIA Blackwell Ultra GPU
Coherent Memory
748 GB unified: 496 GB LPDDR5X + 252 GB HBM3e
AI Performance
20 PetaFLOPS (FP4) / 10 PetaFLOPS (FP8)
Storage & Physical
Storage (Slots)
2× M.2 2280 PCIe 5.0 x4 NVMe (system) + 2× M.2 2280 PCIe 6.0 x4 NVMe (64GB/s, training data path)
Form Factor
Full-tower deskside, NVIDIA DGX Station architecture
Dimensions
24.4 x 59.4 x 29.5 cm
Networking & Expansion
High-Speed Networking
2× 400G QSFP112 ports via NVIDIA ConnectX-8 SuperNIC, plus 1× 10GbE RJ45 management port
PCIe Expansion
PCIe 5.0, 1× double-width x16 slot; supports an optional discrete RTX PRO 6000 Blackwell GPU (96 GB GDDR7)
Power & Thermal
Power Supply
1600 W ATX power supply, Titanium efficiency, standard office outlet
Cooling
Closed-loop liquid cooling, tuned for office-suitable acoustics under sustained AI workloads
Management & Software
BMC
ASPEED AST2600
Remote Management
IPMI 2.0 and Redfish API, data-center-style monitoring and control
GPU Partitioning
Multi-Instance GPU (MIG), up to 7 isolated instances with guaranteed quality of service
Software Stack
Preloaded Linux environment with NVIDIA AI Enterprise tooling, ready for NVIDIA NemoClaw agent deployment
Exxact introduced the Valence DGX Station on January 20, 2026, a deskside system built on the NVIDIA GB300 Grace Blackwell Superchip platform under NVIDIA's DGX Station specification, designed to run trillion-parameter models locally from a single office-friendly tower.
Hardware Specifications
The Valence is built around a single NVIDIA Grace CPU Superchip — 72 Arm Neoverse V2 cores and 496 GB LPDDR5X memory — paired with an NVIDIA Blackwell Ultra GPU with 252 GB HBM3e GPU memory, unified through NVLink-C2C into a coherent memory pool of 748 GB. Storage is split across two tiers: 2× M.2 2280 PCIe 5.0 x4 NVMe for the operating system, and 2× M.2 2280 PCIe 6.0 x4 NVMe (64GB/s) reserved as a dedicated path for training data and model checkpoints, keeping day-to-day system I/O separate from heavy AI workloads. Networking is handled by a dual-port ConnectX-8 400G SuperNIC, complemented by a 10GbE port for out-of-band management.
A double-width PCIe 5.0 x16 slot allows a discrete RTX PRO 6000 Blackwell GPU with 96 GB GDDR7 to be added to the system.
Local AI Server for Business Use
The Valence is designed to function as a shared local AI resource. Multiple employees can submit document processing tasks — contract review, invoice extraction, report drafting, email triage — throughout the working day, with jobs handled centrally on-site instead of routed through external cloud services.
Through NVIDIA NemoClaw, autonomous agents can run continuously on the system, handling recurring tasks such as inbox monitoring or scheduled report generation without requiring a person to initiate each request. This allows the Valence to operate as ongoing AI infrastructure for the business, supporting steady document workloads.
Cost Comparison: On-Premises vs. Cloud AI APIs
For SMBs processing high volumes of documents on a daily basis, the difference between an on-premises system and cloud AI billing compounds significantly over time.
| Cost Category | Cloud AI APIs | Valence (On-Premises) |
|---|---|---|
| Initial hardware investment | None | €94k (one-time) |
| Year 1 operating cost | ~€100k – ~€200k | ~€4k (power & maintenance) |
| Year 2 operating cost | ~€100k – ~€200k | ~€4k (power & maintenance) |
| Year 3 operating cost | ~€100k – ~€200k | ~€4k (power & maintenance) |
| Cumulative 3-year cost | ~€300k – ~€600k | ~€110k |
Because processing takes place entirely on-site, documents and business data are not transmitted to third-party servers. This is a relevant consideration for organizations in legal, financial, or other data-sensitive sectors.
Summary
The Valence DGX Station combines a fixed-cost, deskside-friendly system with the architecture to run large AI models locally, plus the flexibility to add rendering capacity through an optional RTX PRO Blackwell GPU. For SMBs with sustained AI workloads, it offers a lower cost alternative to recurring cloud AI expenditure.
