Large, complex software applications are pushing the performance limitations of today's engineering workstations. Such applications frequently require virtual memory in excess of available physical memory, and performance becomes dominated by paging time. Paging performance is ultimately dependent on the speed of disk I/0, which runs orders of magnitude slower than the maximum performance possible with physical memory and CPU speed alone. Paging performance is directly related to the number of different pages the application touches (i.e., read or write) per unit time. The inverse of this measure can be thought of as a measure of,"locality of reference". When the number of pages touched per unit time exceeds the amount of free physical memory, performance begins to degrade. This project aims at providing a mechanism to increase locality of reference in an existing Common Lisp system by clustering objects in virtual memory with the goal of decreasing the number of pages touched per unit time. Researchers are proposing two components of work: firstly, to prototype tools for reporting useful information on paging and locality of reference; and secondly, to investigate methods for automatically increasing locality of reference during application execution.The potential commercial application as described by the awardee: Common Lisp is particularly well suited for solving very complex software engineering problems, in fields including Scheduling, Process Control, Computer Aided Design, Computer Aided Software Engineering, Artificial Intelligence, and Automatic Program Generation. Improving locality of reference significantly increases the viability of developing and deploying such applications on stock hardware.
