Gaze-Contingent Foveated Rendering

This simulator visualizes gaze-contingent foveated rendering, the technique eye-tracked VR headsets use to cut GPU shading cost by rendering full detail only where the eye is actually looking. A grid of instanced tiles stands in for a headset's render target; each tile's height and brightness encode its local resolution scale, computed live from its retinal eccentricity — the angle between the eye's line of sight to the gaze point and to that tile — using a cone-density-style falloff model. Drag the gaze marker to relocate the high-resolution foveal disk, tune the foveal radius, falloff steepness and peripheral resolution floor, or trigger an auto-saccade sequence, while live readouts report the fraction of native pixels actually rendered and the resulting GPU shading speedup.