Publisher Summary The chapter presents the concept of unformatted databases and parallel architectures, proposed to manipulate textual databases. Databases fall into two general categories—namely, formatted and unformatted structures. Formatted databases are mainly time variant entities and are subject to extensive alteration as well as search operations. Unformatted databases (bibliographic or full-text) are archival in nature and are processed by searching for a pattern or a combination of patterns. The problem of searching large textual databases is addressed in the chapter. To improve the performance of such a lengthy operation, two major directions are discussed: one based on the design of efficient algorithms for pattern matching operations, and the other based on the hardware implementation of the basic pattern matching operations. Both approaches have their own merits and are subject to further research and study. However, the major theme of this chapter is centered around the design of the hardware pattern matcher. Such an emphasis is mainly due to the current advances in technology that have enabled the migration of the software functions into the hardware. Three different schemes of hardware implementation of an efficient term comparator for specialized backend text retrieval architectures are also discussed.