Database management system support for architecture with intelligent memories

By adding logic to the memory, intelligent memories allow the creation of a cost-effective parallel architecture that can address the computational needs of data intensive applications. Currently, database applications are the most predominant data intensive applications because the database instances continue to increase at a rapid rate. This paper introduces database management techniques designed to take advantage of intelligent memory in order to meet the performance needs of current massive database applications. In particular the bitmap indexing, which is a very common indexing in current database management systems, is optimized for intelligent memories. Results suggest that the use of intelligent memories can improve query processing in large databases.