A core idea in computing is combining a small and fast memory with a large but slow memory

Date created: 2022-07-18

If you layer this in multiple steps–a cache for the cache–you have a “hierarchy” of memories, each with a larger but slower capacity. Modern laptops often have six-layer caches.

The CPU capacity has had exponential growth, but memory performance hasn’t. Thus the need for caches.


  • Link to website, bibtex from Zotero or note with book/blog/etc summary