{"ID":2890434,"CreatedAt":"2026-06-01T04:54:23.091178241Z","UpdatedAt":"2026-06-01T04:54:23.091178241Z","DeletedAt":null,"paper_url":"https://arxiv.org/abs/2507.19133","arxiv_id":"2507.19133","title":"3DGauCIM: Accelerating Static/Dynamic 3D Gaussian Splatting via Digital CIM for High Frame Rate Real-Time Edge Rendering","abstract":"Dynamic 3D Gaussian splatting (3DGS) extends static 3DGS to render dynamic scenes, enabling AR/VR applications with moving objects. However, implementing dynamic 3DGS on edge devices faces challenges: (1) Loading all Gaussian parameters from DRAM for frustum culling incurs high energy costs. (2) Increased parameters for dynamic scenes elevate sorting latency and energy consumption. (3) Limited on-chip buffer capacity with higher parameters reduces buffer reuse, causing frequent DRAM access. (4) Dynamic 3DGS operations are not readily compatible with digital compute-in-memory (DCIM). These challenges hinder real-time performance and power efficiency on edge devices, leading to reduced battery life or requiring bulky batteries. To tackle these challenges, we propose algorithm-hardware co-design techniques. At the algorithmic level, we introduce three optimizations: (1) DRAM-access reduction frustum culling to lower DRAM access overhead, (2) Adaptive tile grouping to enhance on-chip buffer reuse, and (3) Adaptive interval initialization Bucket-Bitonic sort to reduce sorting latency. At the hardware level, we present a DCIM-friendly computation flow that is evaluated using the measured data from a 16nm DCIM prototype chip. Our experimental results on Large-Scale Real-World Static/Dynamic Datasets demonstrate the ability to achieve high frame rate real-time rendering exceeding 200 frame per second (FPS) with minimal power consumption, merely 0.28 W for static Large-Scale Real-World scenes and 0.63 W for dynamic Large-Scale Real-World scenes. This work successfully addresses the significant challenges of implementing static/dynamic 3DGS technology on resource-constrained edge devices.","short_abstract":"Dynamic 3D Gaussian splatting (3DGS) extends static 3DGS to render dynamic scenes, enabling AR/VR applications with moving objects. However, implementing dynamic 3DGS on edge devices faces challenges: (1) Loading all Gaussian parameters from DRAM for frustum culling incurs high energy costs. (2) Increased parameters fo...","url_abs":"https://arxiv.org/abs/2507.19133","url_pdf":"https://arxiv.org/pdf/2507.19133v1","authors":"[\"Wei-Hsing Huang\",\"Cheng-Jhih Shih\",\"Jian-Wei Su\",\"Samuel Wade Wang\",\"Vaidehi Garg\",\"Yuyao Kong\",\"Jen-Chun Tien\",\"Nealson Li\",\"Arijit Raychowdhury\",\"Meng-Fan Chang\",\"Yingyan\",\"Lin\",\"Shimeng Yu\"]","published":"2025-07-25T10:16:44Z","proceeding":"cs.AR","tasks":"[\"cs.AR\"]","methods":"[]","has_code":false}