Brev Connect

Brev Connect lets you bring hardware you own into NVIDIA Brev. Connect a supported Linux machine — DGX Spark, DGX Station, or another node — to get SSH access, team sharing, and port forwarding through the Brev platform.

What is Brev Connect?

Brev Connect links your physical hardware to Brev’s management layer. Once registered, your machine appears in the Brev console alongside cloud GPU instances, and your team can connect to it using the same CLI commands.

This is designed for hardware you own and manage — Brev handles connectivity and access control, not the machine’s lifecycle.

How It Works

In the Brev Console, an organization Owner or Admin opens Brev Connect, names the machine, and creates or supplies an unexpired Read & Write personal API key. The generated command installs the Brev CLI and registers the machine with the key’s organization. Run it on the machine you want to connect. See Connect DGX Spark for this workflow and the manual CLI alternative.

Registration performs three steps:

1. Installs NetBird — A mesh VPN agent that provides secure connectivity between your machine and Brev users
2. Profiles hardware — Auto-detects GPUs (via NVML), CPU, RAM, storage, OS, and interconnects like NVLink and PCIe
3. Registers with Brev — Makes the machine visible in your Brev organization

After registration, run brev enable-ssh on the machine, then brev grant-ssh to grant access to your own Brev user or teammates. Authenticate these commands with your API key or a Brev user login. Enabling SSH alone does not grant anyone access. Each person connecting runs brev login on their own computer before brev shell; API key authentication does not set up their user SSH credentials.

Register → NetBird installed → Hardware profiled → Node appears in Brev
    ↓
Enable SSH → Grant user access → Log in on your computer → Connect via brev shell

Brev Connect vs Cloud GPU Instances

Supported Hardware

Brev Connect registration requires Linux, sudo access, and a host that can run the NetBird service with systemd. SSH access also requires a running SSH server. Primary use cases include:

• NVIDIA DGX Spark — Desktop AI workstation
• NVIDIA DGX Station — Deskside AI workstation
• Other Linux nodes — Machines that meet the registration and SSH prerequisites above

Hardware Auto-Detection

During registration, Brev collects the following hardware information when available:

• GPUs — Model, count, and memory via NVML
• CPU — Architecture and logical processor count
• RAM — Total system memory
• Storage — Disk capacity and type
• Interconnects — NVLink and PCIe topology
• OS — Distribution and version
• Product name — e.g., “DGX Spark”, “DGX Station”

This profile is visible in the Brev console and helps teammates understand the machine’s capabilities.

GPU and interconnect detection requires NVIDIA drivers with NVML available. Registration does not install GPU drivers or workload software.

Registration Data

Registration state is stored locally at /etc/brev/device_registration.json. This file is created during brev register and removed during brev deregister.

Next Steps