> ## Documentation Index
> Fetch the complete documentation index at: https://jncpkg.org/llms.txt
> Use this file to discover all available pages before exploring further.

# jnc probe

> Detect NVIDIA Jetson hardware and report system profile information.

# jnc probe

Detects NVIDIA Jetson hardware on the current system and reports a detailed hardware profile. Use this command to verify that your Jetson device is recognized correctly before setting up workspaces, configuring system requirements, or troubleshooting platform-specific issues.

## Usage

```bash theme={null}
jnc probe [OPTIONS]
```

## Options

| Option   | Default | Description                   |
| -------- | ------- | ----------------------------- |
| `--json` | `true`  | Output results in JSON format |

## Output

When Jetson hardware is detected, the command outputs a JSON object containing a `JetsonProfile` with the following fields:

| Field                | Type           | Description                                                                                                                                          | Example              |
| -------------------- | -------------- | ---------------------------------------------------------------------------------------------------------------------------------------------------- | -------------------- |
| `arch`               | String         | CPU architecture of the system                                                                                                                       | `"aarch64"`          |
| `l4t`                | String or null | Linux for Tegra (L4T) version. Queried from the `nvidia-l4t-core` dpkg package on JetPack 6+, or read from `/etc/nv_tegra_release` on older releases | `"36.4.0"`           |
| `jetpack`            | String or null | NVIDIA JetPack SDK version corresponding to the installed L4T release                                                                                | `"6.1"`              |
| `cuda`               | String or null | CUDA driver version. Read from `/proc/driver/nvidia/version`, falling back to `nvidia-smi` if unavailable                                            | `"12.2"`             |
| `module`             | String or null | Jetson module name, read from `/proc/device-tree/model`                                                                                              | `"Jetson Orin Nano"` |
| `compute_capability` | String or null | CUDA compute capability of the GPU                                                                                                                   | `"8.7"`              |

Fields shown as "String or null" will be `null` in the JSON output when the value cannot be determined (for example, when the corresponding file or package is not present).

## Supported Hardware

Currently supports the Jetson Orin Nano with JetPack 6 (L4T R36.x). Support for additional Jetson modules and JetPack versions is coming soon.

## Examples

### Jetson Orin Nano

```bash theme={null}
jnc probe
```

```json theme={null}
{
  "arch": "aarch64",
  "l4t": "36.4.0",
  "jetpack": "6.1",
  "cuda": "12.2",
  "module": "Jetson Orin Nano",
  "compute_capability": "8.7"
}
```

### Non-Jetson System

On a system without Jetson hardware, the command returns a profile with null values for hardware-specific fields:

```bash theme={null}
jnc probe
```

```json theme={null}
{
  "arch": "x86_64",
  "l4t": null,
  "jetpack": null,
  "cuda": null,
  "module": null,
  "compute_capability": null
}
```

## Integration with System Requirements

You can use the L4T version reported by `jnc probe` to configure system requirements in your `jnc.toml`. This ensures that packages requiring a specific L4T release are resolved correctly:

```toml theme={null}
[system-requirements]
l4t = "36.4.0"
```

Run `jnc probe` first to confirm the L4T version on your device, then set the value in `jnc.toml` to match.