I. Introduction
Fuzzy logic is a strong tool for solving problems, which are hard to define mathematically, such as memory allocation in large communities. When using a main server, users are often granted memory according to their actions, as they go along. This is quite prodigal since for each user the system has to decide how much memory may be used with no regularity present. A procedure that will enable a server to derive required memory according to previous knowledge might come in handy in a case like this. In this paper the authors show a new method for taking data regarding CPU use and memory allocation and optimizing it using an optically based fuzzy logic engine. The paper first reviews the optical fuzzy controller that enables the optimization. Then determining the fuzzy logic rules in an optical manner is discussed. Once the setup is fully explained the case of optimizing actual data from the server of Tel Aviv University is shown and results are explained.