---
title: "Register AI Compute with Brev"
publisher: "nvidia"
type: "playbook"
updated: "2026-09-29T17:27:59.727Z"
description: "A shared GPU workspace your team can access from anywhere"
canonical: "https://build.nvidia.com/spark/brev.md"
---

# Basic idea

NVIDIA Brev is an AI development platform that makes GPU environments remotely accessible, shareable, and easy to standardize using preconfigured setups called Launchables.

This walkthrough helps you connect your hardware platform to Brev so it appears as a managed GPU environment. After registration and SSH setup, your hardware becomes remotely accessible and shareable.

# What you'll accomplish

You'll register your hardware platform with Brev and configure SSH access. It will appear as a healthy node in the Brev web UI and CLI, ready to share access and accept workloads.

# What to know before starting

**Required:**

- Familiarity with the Linux command line to run a few setup commands

**Optional:**

- Basic understanding of SSH access and team/org membership in cloud developer portals

# Supported hardware platforms

Use the matrix below to confirm your hardware platform. The same registration workflow applies across supported hardware platforms.

| Hardware platform | OS | Memory  | Multi-node capable hardware |
| :---- | :---- | :---- | :---- |
| **DGX Spark** | DGX OS | 128 GB Unified Memory | — |
| **DGX Station** | DGX OS | Large HBM + Grace DRAM | — |

> [!NOTE]
> Only platforms listed in the Supported hardware platforms table above are covered by this playbook.

# Prerequisites

**Hardware requirements**

- Supported hardware platform — see Supported hardware platforms matrix above
- Hardware platform is powered on, networked, and reachable for local or SSH terminal access

**Software requirements**

- An NVIDIA Brev account — [create an account](https://login.brev.nvidia.com/signin) if you do not have one
- Owner or Admin access to your Brev org
- Administrative (root or sudo) access on the hardware platform to run the registration command
- Network access from the hardware platform to Brev services

# Time & risk

- **Estimated time:** 5–10 MIN
- **Risk level:** Low
- Registration configures the hardware platform for secure remote access without altering existing workloads
- **Rollback:** Run `brev deregister` on the hardware platform (see Cleanup in the **Instructions** tab)
- **Last Updated:** 09/21/2026
- Registration workflow for connecting supported hardware platforms to NVIDIA Brev

## More

- [Instructions](/spark/brev/instructions.md)
- [Troubleshooting](/spark/brev/troubleshooting.md)