Digital Twins on AMD: Building Robotic Simulations Using Edge AI PCs

2026-02-09 · Source: AMD ROCm Blogs · Field: Technology & Digital — Robotics & Autonomous Systems, Artificial Intelligence & Machine Learning · Depth: Intermediate, medium

Summary

AMD's blog post introduces digital twins as a critical tool for robotics, automation, and intelligent systems, enabling virtual validation and data generation before real-world deployment. It highlights the capability of AMD hardware, specifically Ryzen AI MAX, to run high-fidelity physics simulations and parallel environments directly on edge devices. The article defines digital twins, details essential platform components like general-purpose physics engines and high-speed parallelization, and provides a hands-on tutorial using the open-source Genesis robotic simulation platform. This tutorial covers initializing the Genesis backend, creating a simulation scene, adding entities like a Franka Panda arm, building the scene, configuring PD controllers for joint control, executing basic motion planning with inverse kinematics, and scaling simulations to multiple parallel environments, all leveraging AMD's GPU-accelerated stack.

Key takeaway

For Robotics Engineers developing and testing complex systems, integrating digital twins on AMD Ryzen AI MAX with platforms like Genesis can significantly accelerate development cycles. You can validate robot behaviors and generate synthetic datasets in high-fidelity, parallel simulations directly on an edge device, reducing reliance on physical hardware and enabling more proactive maintenance strategies.

Key insights

Digital twins on AMD Ryzen AI MAX enable high-fidelity, parallel robotic simulations directly at the edge.

Principles

• Digital twins reduce risk and accelerate development.
• GPU acceleration is critical for scalable simulations.
• PD controllers ensure realistic robot motion.

Method

The Genesis platform workflow involves initializing the backend, creating a scene with entities, building the scene, configuring PD controllers for DOFs, and then stepping the simulation, with options for IK motion planning and parallel environments.

In practice

• Use Genesis for GPU-accelerated robotics simulation.
• Leverage Ryzen AI MAX for edge AI simulations.
• Employ PD controllers for realistic joint movement.

Topics

• Digital Twins
• Robotic Simulation
• Edge AI
• Genesis Platform
• Parallel Simulation

Code references

• Genesis-Embodied-AI/Genesis
• AMDResearch/aup-learning-cloud
• AMDResearch/aup-learning-cloud
• AMDResearch/Ryzers
• facebookresearch/fairo

Best for: Robotics Engineer, AI Engineer, Machine Learning Engineer

Related on AIssential

• CEO Fireside Chat With Jensen Huang and Pascal Daloz at Dassault Systèmes 3DEXPERIENCE World — AI and virtual twins are converging to reinvent computing infrastructure and industrial processes at unprecedented scales.
• AMD unveils AI accelerators and laptop chips at CES 2026 — AMD's new AI accelerators and laptop chips aim to significantly boost AI performance and expand market reach.
• NVIDIA and SK Hynix: Partnering on Next‑Gen Memory for AI — Strategic partnerships are crucial for scaling AI infrastructure through co-development of advanced memory and AI-driven manufacturing.
• Nvidia, Dassault Systèmes to Build Industrial AI Platform — The partnership combines virtual twin technology with AI to create science-validated industrial world models and AI companions.
• Edge-to-Cloud Robotics with AMD ROCm: From Data Collection to Real-Time Inference — AMD's edge-to-cloud robotics solution integrates LeRobot, VLA models, and Imitation Learning on AMD hardware for efficient robot training and deployment.
• Cheers to AI: ADAM Robot Bartender Makes Drinks at Vegas Golden Knights Game — Robots trained in simulation with synthetic data can perform complex real-world tasks with high precision and low latency.

Open in AIssential →

Editorial summary, takeaway, and curation by AIssential. Original article published by AMD ROCm Blogs.