Skip to main content
AI Features

Workstation Setup Guide

This guide provides comprehensive instructions for setting up a high-performance workstation optimized for Physical AI and humanoid robotics development. Follow these steps carefully to ensure your environment meets the demanding requirements of modern robotics simulation and AI development.

Overview​

Physical AI development requires substantial computational resources for:

  1. Real-time physics simulation (Gazebo, Isaac Sim)
  2. AI model training and inference
  3. Computer vision processing
  4. 3D rendering and visualization
  5. Multi-sensor data processing

Hardware Requirements​

Minimum Specifications​

Processor: Intel Core i7 (13th Gen) or AMD Ryzen 9

  • 8+ cores, 16+ threads preferred
  • Base clock 3.0+ GHz
  • Support for AVX2 instruction set

Memory: 32GB DDR5 (minimum)

  • 64GB DDR5 recommended for complex simulations
  • Dual-channel or quad-channel configuration

GPU: NVIDIA RTX 4070 Ti (12GB VRAM) - minimum

  • RTX 4090 (24GB VRAM) recommended
  • Must support CUDA 11.8+ and ray tracing
  • Multiple GPU support for advanced workloads

Storage: 1TB NVMe SSD (minimum)

  • 2TB NVMe SSD recommended
  • Additional 4TB HDD for data storage
  • USB-C/Thunderbolt external storage support

Operating System: Ubuntu 22.04 LTS (mandatory)

  • Native installation preferred
  • Dual-boot configuration acceptable
  • Virtual machines not supported for real-time workloads

Development Workstation (Premium)​

CPU: Intel Core i9-13900K or AMD Ryzen 9 7950X
GPU: NVIDIA RTX 4090 (24GB VRAM)
RAM: 64GB DDR5-5600 (4x16GB)
Storage: 2TB NVMe SSD + 4TB HDD
Cooling: Liquid cooling system
Power: 1000W 80+ Platinum PSU
Case: Full tower with excellent airflow

Research Workstation (Intermediate)​

CPU: Intel Core i7-13700K or AMD Ryzen 7 7700X
GPU: NVIDIA RTX 4080 (16GB VRAM)
RAM: 32GB DDR5-5200 (2x16GB)
Storage: 1TB NVMe SSD + 2TB HDD
Cooling: High-end air cooling
Power: 850W 80+ Gold PSU
Case: Mid-tower with good ventilation

Ubuntu 22.04 LTS Installation​

Pre-Installation Checklist​

  1. Backup Data: Ensure all important data is backed up
  2. Verify Compatibility: Check hardware compatibility with Ubuntu 22.04
  3. Create Installation Media: USB drive (8GB+ minimum)
  4. Network Access: Ensure internet connectivity for updates

Installation Process​

  1. Download Ubuntu 22.04 LTS

    # Verify download integrity
    sha256sum ubuntu-22.04.3-desktop-amd64.iso

  2. Create Bootable USB

    # Using Rufus (Windows) or dd (Linux/Mac)
    dd if=ubuntu-22.04.3-desktop-amd64.iso of=/dev/sdX bs=4M status=progress

  3. Boot from USB

    • Enter BIOS/UEFI setup (F2, F12, or Del during boot)
    • Disable Secure Boot (temporarily)
    • Set USB drive as first boot device
    • Enable virtualization (VT-x/AMD-V)

  4. Partition Scheme (Recommended)

    /boot/efi     512MB   EFI System Partition
    /boot          2GB     Ext4 (boot files)
    swap          8GB      Swap space (equal to RAM for hibernation)
    /            100GB    Ext4 (root filesystem)
    /home         500GB   Ext4 (user data)
    /var          200GB   Ext4 (applications, logs)
    /tmp          50GB    Ext4 (temporary files)

Post-Installation Configuration​

  1. Update System

    sudo apt update && sudo apt upgrade -y
    sudo apt install -y curl wget git vim htop

  2. Enable Additional Drivers

    # Install proprietary NVIDIA drivers
    ubuntu-drivers autoinstall
    sudo reboot

  3. Configure SSH Access

    sudo apt install -y openssh-server
    sudo systemctl enable ssh
    sudo ufw allow ssh

NVIDIA Driver and CUDA Setup​

Driver Installation​

  1. Verify GPU Detection

    lspci | grep -i nvidia
    nvidia-smi

  2. Install Latest NVIDIA Drivers

    # Add NVIDIA repository
    wget https://developer.download.nvidia.com/compute/cuda/repos/ubuntu2204/x86_64/cuda-keyring_1.1-1_all.deb
    sudo dpkg -i cuda-keyring_1.1-1_all.deb
    sudo apt-get update

    # Install driver and CUDA toolkit
    sudo apt-get -y install cuda-toolkit-12-2

  3. Configure Environment

    echo 'export PATH=/usr/local/cuda/bin${PATH:+:${PATH}}' >> ~/.bashrc
    echo 'export LD_LIBRARY_PATH=/usr/local/cuda/lib64${LD_LIBRARY_PATH:+:${LD_LIBRARY_PATH}}' >> ~/.bashrc
    source ~/.bashrc

CUDA Verification​

# Verify CUDA installation
nvcc --version

# Test with sample compilation
cd /usr/local/cuda/samples/1_Utilities/deviceQuery
make
./deviceQuery

ROS 2 Humble Installation​

Setup Instructions​

  1. Set Locale

    sudo apt install -y locales
    sudo locale-gen en_US en_US.UTF-8
    sudo update-locale LC_ALL=en_US.UTF-8 LANG=en_US.UTF-8
    export LANG=en_US.UTF-8

  2. Add ROS 2 Repository

    sudo apt install -y software-properties-common
    sudo add-apt-repository universe
    curl -sSL https://raw.githubusercontent.com/ros/rosdistro/master/ros.key -o /usr/share/keyrings/ros-archive-keyring.gpg
    echo "deb [arch=$(dpkg --print-architecture) signed-by=/usr/share/keyrings/ros-archive-keyring.gpg] http://packages.ros.org/ros2/ubuntu $(. /etc/os-release && echo $UBUNTU_CODENAME) main" | sudo tee /etc/apt/sources.list.d/ros2.list > /dev/null

  3. Install ROS 2 Humble

    sudo apt update
    sudo apt install -y ros-humble-desktop

  4. Development Tools

    sudo apt install -y python3-pip python3-venv
    pip install --user pip-upgrade setuptools
    pip install --user rosdep
    sudo rosdep init
    rosdep update

Environment Setup​

# Add ROS 2 to bash session
echo "source /opt/ros/humble/setup.bash" >> ~/.bashrc
source ~/.bashrc

# Verify installation
ros2 doctor

Development Environment Configuration​

Python Environment​

  1. Install Python 3.10

    sudo apt install -y python3.10 python3.10-venv python3.10-dev

  2. Create Virtual Environment

    python3.10 -m venv ~/ros2_ws/env
    source ~/ros2_ws/env/bin/activate
    pip install --upgrade pip

  3. Install Essential Packages

    pip install numpy scipy matplotlib opencv-python
    pip install torch torchvision torchaudio --index-url https://download.pytorch.org/whl/cu118
    pip install tensorflow-gpu
    pip install jupyter notebook pandas scikit-learn

ROS 2 Workspace Setup​

# Create workspace
mkdir -p ~/ros2_ws/src
cd ~/ros2_ws/src

# Clone essential packages
git clone https://github.com/ros-geometry/geometry2.git
git clone https://github.com/ros-perception/vision_opencv.git
git clone https://github.com/ros-drivers/usb_cam.git

# Build workspace
cd ~/ros2_ws
colcon build --symlink-install

# Source workspace
echo "source ~/ros2_ws/install/setup.bash" >> ~/.bashrc
source ~/.bashrc

Performance Optimization​

System Tuning​

  1. CPU Governor Settings

    # Set to performance mode
    sudo cpupower frequency-set -g performance

    # Make permanent
    echo 'GOVERNOR="performance"' | sudo tee /etc/default/cpufrequtils

  2. Memory Management

    # Configure swappiness
    echo 'vm.swappiness=10' | sudo tee -a /etc/sysctl.conf

    # Configure dirty pages
    echo 'vm.dirty_ratio=15' | sudo tee -a /etc/sysctl.conf
    echo 'vm.dirty_background_ratio=5' | sudo tee -a /etc/sysctl.conf

  3. GPU Performance

    # Set GPU performance mode
    sudo nvidia-smi -pm 1
    sudo nvidia-smi -ac 877,1215

    # Create persistence script
    echo '#!/bin/bash
    sudo nvidia-smi -pm 1
    sudo nvidia-smi -ac 877,1215' | sudo tee /usr/local/bin/gpu-performance
    sudo chmod +x /usr/local/bin/gpu-performance

File System Optimization​

  1. SSD Optimization

    # Enable fstrim timer
    sudo systemctl enable fstrim.timer
    sudo systemctl start fstrim.timer

    # Configure I/O scheduler
    echo 'mq-deadline' | sudo tee /sys/block/sdX/queue/scheduler

  2. Tmpfs Configuration

    # Add to /etc/fstab for temporary files in RAM
    echo 'tmpfs /tmp tmpfs defaults,noatime,mode=1777 0 0' | sudo tee -a /etc/fstab
    echo 'tmpfs /var/tmp tmpfs defaults,noatime,mode=1777 0 0' | sudo tee -a /etc/fstab

Verification and Testing​

System Performance Test​

# CPU performance
sysbench cpu --cpu-max-prime=20000 run

# Memory performance
sysbench memory --memory-block-size=1K --memory-total-size=10G run

# GPU benchmark
nvidia-smi dmon -s u -d 1

# ROS 2 latency test
ros2 run rclcpp_examples latency_test

ROS 2 Functionality Test​

# Test basic ROS 2 functionality
ros2 run demo_nodes_cpp talker &
ros2 run demo_nodes_py listener

# Test gazebo simulation
ros2 launch gazebo_ros empty_world.launch.py

# Verify GPU acceleration
ros2 run rviz2 rviz2  # Should open with hardware acceleration

Troubleshooting​

Common Issues​

  1. NVIDIA Driver Issues

    # Reinstall drivers
    sudo apt purge nvidia-*
    sudo ubuntu-drivers autoinstall
    sudo reboot

  2. ROS 2 Environment Issues

    # Reset environment
    unset ROS_DOMAIN_ID
    unset ROS_LOCALHOST_ONLY
    source /opt/ros/humble/setup.bash

  3. Performance Issues

    # Check system resources
    htop
    iotop
    nvidia-smi dmon

    # Check thermals
    sensors
    sudo watch -n 1 nvidia-smi

Getting Help​

Next Steps​

After completing this workstation setup:

  1. GPU Optimization Guide: Proceed to GPU Optimization for advanced GPU tuning
  2. Cloud Setup: Consider Cloud Alternatives for additional compute resources
  3. Hardware Procurement: Review hardware procurement options for robot components

Validation Checklist​

  • Ubuntu 22.04 LTS installed and updated
  • NVIDIA drivers installed and functioning
  • CUDA toolkit installed and verified
  • ROS 2 Humble installed with development tools
  • Python virtual environment configured
  • ROS 2 workspace created and built
  • System performance optimizations applied
  • All verification tests passing
  • Backup and recovery procedures documented

Estimated Setup Time: 4-6 hours Difficulty Level: Intermediate Prerequisites: Basic Linux administration skills Support: Available through course forums and office hours

For cloud-based alternatives or budget-conscious options, see our Cloud Alternatives guide.