Vorion: A RISC-V GPU with Hardware-Accelerated 3D Gaussian Rendering and Training

2025-11-20

AI Summary

To address the high computational overhead and deployment challenges of 3D Gaussian Splatting (3DGS) in real-time rendering on edge devices and workstation-scale 4D reconstruction, this paper presents the first RISC-V-based GPU prototype tailored for 3DGS acceleration. Methodologically, we propose a Gaussian-pixel hybrid dataflow architecture and a z-tiling parallel optimization strategy, enabling seamless integration into standard graphics pipelines with minimal modifications. Implemented in TSMC 16nm technology, our scalable GPGPU architecture incorporates dedicated 3DGS hardware modules. Experimental results demonstrate real-time rendering at 19 FPS on a single chip and a training iteration throughput of 38.6 iterations/second under a 16-rasterizer configuration. To the best of our knowledge, this is the first end-to-end hardware acceleration framework for 3D neural rendering and 4D reconstruction on a RISC-V architecture, delivering an efficient, scalable hardware foundation for intelligent edge vision systems.

Technology Category

• Computer Vision: 3D Computer Vision
• Machine Learning: Hardware-aware ML
• Cognitive Modeling & Cognitive Systems: Neural Spike Coding

Application Category

• Search and Retrieval-Augmented AI: Retrieval-Augmented Generation (RAG) and multi-modal RAG
• Economics, Online Markets and Human Computation: Architectures and workflows that use LLMs for crowd work
• Graph Algorithms and Modeling for the Web: Graph neural networks and deep learning approaches for Web-related graphs

Abstract

3D Gaussian Splatting (3DGS) has recently emerged as a foundational technique for real-time neural rendering, 3D scene generation, volumetric video (4D) capture. However, its rendering and training impose massive computation, making real-time rendering on edge devices and real-time 4D reconstruction on workstations currently infeasible. Given its fixed-function nature and similarity with traditional rasterization, 3DGS presents a strong case for dedicated hardware in the graphics pipeline of next-generation GPUs. This work, Vorion, presents the first GPGPU prototype with hardware-accelerated 3DGS rendering and training. Vorion features scalable architecture, minimal hardware change to traditional rasterizers, z-tiling to increase parallelism, and Gaussian/pixel-centric hybrid dataflow. We prototype the minimal system (8 SIMT cores, 2 Gaussian rasterizer) using TSMC 16nm FinFET technology, which achieves 19 FPS for rendering. The scaled design with 16 rasterizers achieves 38.6 iterations/s for training.

Problem

• Accelerating 3D Gaussian Splatting rendering and training for real-time performance
• Enabling real-time neural rendering on edge devices and workstations
• Designing dedicated GPU hardware for efficient 3DGS computation

Innovation

• RISC-V GPU with hardware-accelerated 3D Gaussian rendering
• Scalable architecture with minimal changes to traditional rasterizers
• Hybrid dataflow combining Gaussian and pixel-centric approaches