29‐1: Common Triangle Optimizations and Coarse Pixel Shading for Foveated Rendering Acceleration

Abstract Improved foveated rendering performance for head‐mounted displays (HMDs) requires re‐architecting a multi‐view render pipeline. This pipeline must manage common triangle culling/clipping operations across views while reordering triangle vs. view processing and varying the rasterization sampling rate. This paper explores render pipelines for HMD rendering improvement.