3D Gaussian Splatting (3DGS) is a breakthrough in 3D reconstruction using 3D Gaussians. However, even on high-end GPUs like the NVIDIA A100, reconstructing complex scenes remains time-consuming, taking over 15 minutes. The main bottleneck is $\alpha$-computation, which accounts for 71.25% of training workload, yet 93.03% of it is invalid due to the localized influence of Gaussians. To address this issue, we propose GauPHP, an accelerator for 3DGS training with Gaussian–Pixel hybrid parallelism.

At the software level, we introduce a hybrid parallel workflow that breaks the limitation of conventional pixel-only parallelism through two key techniques: Center-Pixel Gaussian Culling (CPGC), which eliminates invalid Gaussians early, and Seed-Driven Gaussian Region Exploration (SDGRE), which reduces invalid computation for partially valid Gaussians by selectively exploring valid regions. Overall, the workflow significantly reduces $\alpha$ computations, lowering the workload to 17.99%. At the hardware level, GauPHP decouples $\alpha$-computation and $\alpha$-blending into GUnits and PUnits, organized in a 2D mesh-based NoC that supports asynchronous execution and efficient data routing. We further boost performance via Gaussian/Pixel load balancing and tiled SSIM based pipelining. Evaluation results that GauPHP achieves 19.63$\times$, 14.86$\times$, 15.42$\times$, 2.98$\times$ and 2.63$\times$ speedup, and 78.62$\times$, 63.00$\times$, 61.72$\times$, 3.89$\times$ and 3.22$\times$ energy saving, compared to A100, GSCore, GBU, GSArch, and GauSPU, respectively, with negligible image quality loss.