The objective of this proposal is to strengthen the Biomedical Computing Resource at Washington University by developing the capability for designing and building specialized computing systems using Very Large Scale Integrated circuit and system technology (VLSI). We intend to apply this capability to the design of uniquely powerful and efficient specialized computing systems, in collaboration with biomedical researchers who are unable to solve their problems with conventional computing approaches. Our approach is to extend the system design and construction methodology of macromodules to take advantage of technical advances in commercially available VLSI parts, such microprocessors and memories, and improvements in the ability to design custom integrated circuits and sub-systems made possible be recent developments in computer design aids and in methods for fabricating economical prototypes and small quantities of integrated circuit designs. Our system design methodology is based upon the use of a rigorously defined concept of modularity, realized in the form of asynchronous modules that permit the easy design of highly parallel computing structures with distributed control. Our approach depends upon reduction of the cost and time required to design and realize specialized systems through the use of formal specification and design methods supported by extensive computer aids to the design process. We intend to make maximum use of design aid programs available from other universities, and to focus our own theory and design aid development efforts on techniques for asynchronous design. Our initial application studies will be in research areas in which strong collaborations already exist within the Washington University community. These include Modecular Modelling and Drug Design, Database Systems for Clinical Research, Radiation Treatment Planning, Time of Flight Tomography, Cochlear Modelling, and Ultrasonic Tissue Characterization. We expect that these studies will lead to national collaborative programs such as those previously undertaken by the Washington University Resource.
