Worley noise is a noise function, using a point-based approach to generate a cellular pattern. It’s appearance is related to Voronoi cells: instead of subdividing the space into discrete cells, Worley noise uses the distance to the nearest point as the output that can be plugged into a number of different functions.
In the animation above, there are four separate output functions that each use an identical Worley noise function as an input. By using the same noise function as an input to different output functions, we can create layered patterns that build on each other while remaining independent, meaning they can be calculated quickly and in parallel.