Item 18: Amortize the cost of expected computations.(More Effective C++)

There is a common theme running through this Item, and that's that greater speed can often be purchased at a cost of increased memory usage. Keeping track of running minima, maxima, and averages requires extra space, but it saves time. Caching results necessitates greater memory usage but reduces the time needed to regenerate the results once they've been cached. Prefetching demands a place to put the things that are prefetched, but it reduces the time needed to access those things. The story is as old as Computer Science: you can often trade space for time. (Not always, however. Using larger objects means fewer fit on a virtual memory or cache page. In rare cases, making objects bigger reduces the performance of your software, because your paging activity increases, your cache hit rate decreases, or both.