🚀 Ubuntu Auto Installation Scripts

A complete set of installation scripts to set up a GPU-accelerated AI & robotics development environment on Ubuntu 22.04. This includes NVIDIA drivers, CUDA, PyTorch, Docker, ROS 2, Miniconda, and essential development tools like Chrome, VSCode, and more.

📋 Supported System

• OS: Ubuntu 22.04 LTS (Jammy)
• GPU: NVIDIA (Tested on RTX 4090 with Driver)
• Shell: Bash
• Network: Internet connection required

---

📦 Version Compatibility Summary

Component	Version
OS	Ubuntu 22.04
GPU Driver	NVIDIA auto version
CUDA Toolkit	12.2 (driver-only install)
PyTorch	2.2.2 (CUDA 12.1 pip build)
Python	3.10 (via Miniconda)
ROS2	Humble (desktop full version)

---

📦 Included Installation Scripts

Category	Script	Description
GPU Driver / CUDA	nvidia-driver_install.sh	Install NVIDIA driver (autoinstall)
	cuda_install.sh	Install CUDA Toolkit 12.2
	torch_install.sh	Install PyTorch with CUDA 12.1 support
Docker	Docker_install.sh	Install Docker CE and Compose
	nvidia_docker_toolkit_install.sh	NVIDIA Container Toolkit for GPU Docker
Python Env	miniconda_install.sh	Install Miniconda + disable base env
ROS2	ROS2-humble_install.sh	Install ROS 2 Humble (desktop full)
Development Tools	vscode_install.sh	Install Visual Studio Code
	chrome_install.sh	Install Google Chrome
	cursor_install.sh	Install Cursor editor
Utilities	utils_install.sh	Install Snap + Slack, Discord, Zoom, TeamViewer

---

⚙️ How to Use

1. Clone this repository

git clone https://github.com/yourname/ubuntu-auto-installation.git
cd ubuntu-auto-installation

2. Make scripts executable

chmod +x *.sh
chmod +x scripts/*.sh

3. Run the master install script

# Default installation (auto-selects NVIDIA driver)
bash install_all.sh

# Or specify versions
bash install_all.sh --nvidia-driver 535 --cuda 12.2 --torch 2.2.2 --torch-cuda 12.1

The install order and dependencies are automatically managed.

Installation Order

1. NVIDIA Driver → CUDA → PyTorch
2. Docker + NVIDIA Container Toolkit
3. Miniconda
4. ROS 2 Humble
5. Chrome, VSCode, Cursor
6. Utility Tools (CLI + Snap-based)

After Installation

To finalize the environment:

source ~/.bashrc

Then check:

nvidia-smi
conda --version
docker --version
ros2 --version

ROS 2 Test Example

Open two terminals:

# Terminal 1
ros2 run demo_nodes_cpp talker

# Terminal 2
ros2 run demo_nodes_cpp listener

⚠️ If the listener does not respond, try setting the same domain ID:

export ROS_DOMAIN_ID=10

Project Structure

ubuntu-auto-setup/
├── scripts/
│   ├── chrome_install.sh
│   ├── cuda_install.sh
│   ├── cursor_install.sh
│   ├── Docker_install.sh
│   ├── miniconda_install.sh
│   ├── nvidia-driver_install.sh
│   ├── nvidia_docker_toolkit_install.sh
│   ├── ROS2-humble_install.sh
│   ├── torch_install.sh
│   ├── utils_install.sh
│   ├── vscode_install.sh
├── install_all.sh
└── README.md

🙋 Contribution

PRs and issues are welcome!