Software Rasterization of 2 Billion Points in Real Time

The accelerated collection of detailed real-world 3D data in the form ever-larger point clouds is sparking a demand for novel visualization techniques that are capable rendering billions primitives real-time. We propose software rasterization pipeline up to two billion points real-time (60 FPS) on commodity hardware. Improvements over state art achieved by batching points, enabling number batch-level optimizations before rasterizing them within same pass. These include frustum culling, level-of-detail (LOD) rendering, and choosing appropriate coordinate precision given batch directly compute workgroup. Adaptive precision, conjunction with visibility buffers, reduces required majority just four bytes, making our approach several times faster than bandwidth-limited art. Furthermore, support LOD makes suitable arbitrarily large clouds, meet elevated performance demands virtual reality applications.