TeGo Station

Compute, models and software. Delivered together.

GB10 compute, preinstalled TeGo and local models in one appliance for enterprise deployment.

Book a consultation
TeGo Station
Grace Blackwell Superchip
GB10
LPDDR5x shared CPU / GPU memory
128 GB
M.2 NVMe SSD local storage
2 TB

What’s included

GB10 compute hardware

GB10 appliances with 128 GB of shared memory, a 2 TB SSD and power accessories per unit, supporting multi-appliance deployment.

Predeployed local models

Predeployed local models and inference services, configured and validated for delivery.

Preinstalled TeGo

The digital employee platform with organization, account, permission and business template configuration.

Setup and support

System setup, scenario validation, administrator handover and after-sales support.

Where it fits

Keep sensitive work within your network

Deploy local models and storage within the enterprise network to process customer records, internal documents and other sensitive business materials. Bring the model services and runtime into your own environment so digital employees can work with internal business content.

  • Customer records and internal documents
  • Local models and storage deployed together
  • Delivery aligned with your internal network

Connect business execution with organizational control

Bring digital employees into your organization’s management practices for finance, government and other work with defined access and execution requirements. Configure user access, execution authorization and audit records around the way your organization operates.

  • Access configured for your organization
  • Authorization defines the scope of execution
  • Audit records retain the operation history

Compute, models and applications, ready together

Start with business applications even without an existing local model environment. Compute hardware, predeployed models, inference services and TeGo arrive together, with system setup, scenario validation and administrator handover to support ongoing use and operations.

  • GB10 compute and local storage
  • Predeployed models and inference services
  • Platform installation, setup and support

Expand compute as your business grows

As adoption moves from a few tasks to more people and parallel workloads, compute needs grow too. Connect multiple appliances and add devices to expand compute and concurrent processing capacity, planning each expansion around actual business demand.

  • Multiple appliances deployed together
  • Add devices as workloads grow
  • Expand compute and concurrent processing

Hardware specifications

Compute chip
NVIDIA GB10 Grace Blackwell, 20-core Arm CPU
Unified system memory
128 GB LPDDR5x, shared by CPU and GPU
Local storage
2 TB M.2 NVMe SSD with self-encryption
Wired networking
1 RJ-45 10 GbE port, ConnectX-7 Smart NIC
Wireless connectivity
Wi-Fi 7, Bluetooth 5.4
USB and display
4 USB4 Type-C ports, 1 HDMI 2.1a port
Dimensions and weight
150 × 150 × 50.5 mm, 1.2 kg
Base system
Preinstalled NVIDIA DGX OS with support for the NVIDIA AI software stack and CUDA ecosystem
QR code to join our WeChat group QR code to follow our WeChat official account