Accelerating 3D Gaussian Splatting using Tensor Cores

2026-05-18

AI Summary

This work addresses the computational inefficiency and high latency of the rasterization stage in 3D Gaussian Splatting (3DGS), which fails to leverage the Tensor Core capabilities of modern GPUs. The authors propose the first Tensor Core–accelerated approach for 3DGS by reformulating rasterization as FP16-precision matrix operations through a tensorized rasterization framework. They further introduce a cross-tile grouping strategy to enhance Gaussian data reuse and computational density. This method overcomes the limitations of conventional scalar and irregular execution patterns, achieving a 1.65× end-to-end rendering speedup with negligible loss in image quality.

Abstract

3D Gaussian Splatting (3DGS) has become a leading technique for real-time neural rendering and 3D scene reconstruction, but its rendering cost remains too high for many latency-sensitive scenarios. In particular, the rasterization stage in 3DGS dominates end-to-end rendering time, during which the renderer repeatedly evaluates each Gaussian's contribution to each covered pixel, making this stage compute-bound. At the same time, modern GPUs provide high-throughput Tensor Cores for low-precision matrix operations, yet existing 3DGS systems execute rasterization entirely on CUDA cores and leave Tensor Cores idle. We find that 3DGS rendering can be executed in FP16 with negligible quality degradation, suggesting a promising opportunity for Tensor Core acceleration. However, exploiting Tensor Cores for 3DGS is non-trivial because rasterization does not naturally match their execution model. Existing 3DGS rasterization is expressed as irregular per-pixel scalar operations, whereas Tensor Cores require dense, regular, and reuse-rich matrix workloads. Moreover, conventional tile-by-tile execution fails to exploit Gaussian reuse across neighboring tiles, resulting in repeated data loading and thus high data movement overhead. To this end, we present TensorGS, a 3DGS acceleration framework using Tensor Cores. TensorGS tensorizes the dominant rasterization computation into Tensor-Core-compatible matrix operations and introduces cross-tile grouping to improve Gaussian reuse, amortize overhead, and increase Tensor Core utilization. Experimental results show that TensorGS improves end-to-end rendering performance by 1.65$\times$ while preserving image quality.

Problem

• 3D Gaussian Splatting
• rasterization
• Tensor Cores
• real-time rendering
• performance bottleneck

Innovation

• Tensor Cores
• 3D Gaussian Splatting
• rasterization acceleration
• FP16 computation
• cross-tile reuse