We present a framework for developing High Performance Engineering Information Systems (EIS) based on Incremental Computation Models. These models utilize results of previous computations stored in persistent cache structures by merging them with new results obtained from computations performed on the incremental changes of the database. Because computation is performed on small increments rather than the whole database, incremental computation models achieve performances that are far beyond the reach of ordinary models. High performance EIS will be able to support engineering knowledge bases of significant size, explicit models of interdependent knowledge, optimization techniques for deductive queries, and efficient methods for organizing and accessing large numbers of engineering facts and rules. The persistent cache structures of the incremental models provide an innovative self-improving perforrmance capability for computation intensive EIS.