NeoPhysIx: An Ultra Fast 3D Physical Simulator as Development Tool for AI Algorithms

TL;DR

NeoPhysIx is a highly optimized 3D physical simulator that dramatically accelerates AI training by simplifying physics calculations, enabling real-world scenario simulation speeds exceeding 1000x, thus facilitating faster AI development.

Contribution

The paper introduces NeoPhysIx, a novel ultra-fast 3D physics simulator that overcomes parallelization limitations to achieve unprecedented speedups for AI training.

Findings

Achieves over 1000x speedup compared to real-time simulation.

Successfully trains a legged robot with 18 degrees of freedom in 9 hours.

Demonstrates effective use in AI algorithm development and testing.

Abstract

Traditional AI algorithms, such as Genetic Programming and Reinforcement Learning, often require extensive computational resources to simulate real-world physical scenarios effectively. While advancements in multi-core processing have been made, the inherent limitations of parallelizing rigid body dynamics lead to significant communication overheads, hindering substantial performance gains for simple simulations. This paper introduces NeoPhysIx, a novel 3D physical simulator designed to overcome these challenges. By adopting innovative simulation paradigms and focusing on essential algorithmic elements, NeoPhysIx achieves unprecedented speedups exceeding 1000x compared to real-time. This acceleration is realized through strategic simplifications, including point cloud collision detection, joint angle determination, and friction force estimation. The efficacy of NeoPhysIx is demonstrated through its application in training a legged robot with 18 degrees of freedom and six sensors, controlled by an evolved genetic program. Remarkably, simulating half a year of robot lifetime within a mere 9 hours on a single core of a standard mid-range CPU highlights the significant efficiency gains offered by NeoPhysIx. This breakthrough paves the way for accelerated AI development and training in physically-grounded domains.