If you render two triangles per pixel, the image can be indistinguishable from one rendered with thousands of triangles per pixel, but rendering less than that will show visible artifacts.

Rendering objects at extreme close-up and far away distances requires generating the right amount of geometry to be indistinguishable from reality without excess.

The human eye's focus and attention mechanism determines how much detail should be shown in a scene, requiring high resolution for areas of focus while allowing less detail elsewhere.

Rendering a perfect single frame is easier than rendering a series of frames in motion that look perfect, as earlier algorithms often resulted in visible transitions between different triangle counts.

One breakthrough in graphics was designing a pipeline that avoids the triangle rasterization hardware in GPUs, allowing for better results through pixel shader calculations and ray tracing.

The optimization of rendering in Unreal Engine was a significant challenge, focusing on reducing the cost of rendering pixels to improve performance.

The challenge in CT imaging is that the data collected is often insufficient for high-resolution images, leading to noise in the measurements.

Nanite technology in Unreal Engine minimizes visible transitions in geometry, making them statistically present but not perceptible to the viewer.

We had to be very paranoid about scaling laws, ensuring that any wins from new features were persistent across different scales.