PyBatchRender: A Python Library for Batched 3D Rendering at Up to One Million FPS

2026-01-03

arXiv.org

AI Summary

This work addresses the challenge that reinforcement learning from pixel inputs is often hindered by the performance and complexity of 3D rendering environments, which struggle to simultaneously achieve high frame rates and ease of use. To overcome this limitation, the authors present a high-performance, pure-Python batched 3D rendering library built on the Panda3D engine, specifically designed for reinforcement learning. The library enables flexible and rapid creation of custom scenes while maintaining a simple Python interface. Through batched rendering optimizations, it achieves up to a 1000× speedup over conventional approaches, reaching over one million frames per second in simple scenarios. This dramatic acceleration significantly enhances training efficiency, and entire scenes can be deployed with only a few dozen lines of code.

Technology Category

• Machine Learning: Reinforcement Learning
• Computer Vision: 3D Computer Vision
• Natural Language Processing: Prompt Engineering / Prompting

Application Category

• Economics, Online Markets and Human Computation: Cost models of using LLMs in production systems
• Search and Retrieval-Augmented AI: Web learning to rank, online learning, and counterfactual learning for ranking
• Responsible Web: Data and user privacy-enhancing technologies for the Web

Abstract

Reinforcement learning from pixels is often bottlenecked by the performance and complexity of 3D rendered environments. Researchers face a trade-off between high-speed, low-level engines and slower, more accessible Python frameworks. To address this, we introduce PyBatchRender, a Python library for high-throughput, batched 3D rendering that achieves over 1 million FPS on simple scenes. Built on the Panda3D game engine, it utilizes its mature ecosystem while enhancing performance through optimized batched rendering for up to 1000X speedups. Designed as a physics-agnostic renderer for reinforcement learning from pixels, PyBatchRender offers greater flexibility than dedicated libraries, simpler setup than typical game-engine wrappers, and speeds rivaling state-of-the-art C++ engines like Madrona. Users can create custom scenes entirely in Python with tens of lines of code, enabling rapid prototyping for scalable AI training. Open-source and easy to integrate, it serves to democratize high-performance 3D simulation for researchers and developers. The library is available at https://github.com/dolphin-in-a-coma/PyBatchRender.

Problem

• reinforcement learning from pixels
• 3D rendering
• performance bottleneck
• simulation speed
• rendering complexity

Innovation

• batched rendering
• high-throughput 3D simulation
• reinforcement learning from pixels
• Python-based renderer
• Panda3D