Project Description

Status of Project

Quotable Quotes

The run-time of a program is sometimes dominated not by the number of instructions it performs but by memory effects -- cache misses, the number page swaps, etc. Some algorithmic research has focused on minimizing some particular memory effect -- say, minimizing the number of page faults. Can we minimize *all* memory effects simultaneously? Is this too much to ask for? We consider both theoretical and experimental aspects of this question.