You can use Maya threading to preload your cached deformation data into your system’s memory for faster playback. When a geometry cache’s Multi Thread attribute is on, one thread of Maya will load the cache’s data into memory, and another thread of Maya will playback the cached deformations. If the geometry cache’s Queue Size (measured in frames) is smaller than the cache, then the Maya threads will work together to preload the specified cache frames and playback those frames in the scene view. If the geometry cache’s Queue Size is greater than the size of the cache, then the entire geometry cache will be preloaded into your system’s memory. For best results, use a queue size larger than the frame range of your geometry cache and ensure that your system has sufficient memory (RAM) for the size of your geometry cache.
See cacheFile for information on threading-related geometry cache attributes.