SYNCHRONOUS CIRCUIT OPTIMIZATION VIA MULTIDIMENSIONAL RETIMING

Time-critical sections of multidimensional applications, such as image processing and computational fluid dynamics are in general iterative or recursive. Most of these applications require each iteration to be executed under a specific time constraint associated with the data inp ut rate. The design of circuits dedicated to perform such repetitive t asks depend on optimization techniques to achieve the desired executio n time. The retiming technique is one of these optimization tools; how ever the traditional retiming deals only with one dimension of the pro blem and has lower bound constraints in the execution time due to char acteristics of the initial design, This paper presents a novel optimiz ation technique based on the application of a multidimensional retimin g. Multidimensional retiming improves the circuitry performance by ins erting a fixed number of registers, which is independent of the size o f the problem, into the circuit paths, and restructuring the memory el ements in a legal way. This technique guarantees that all functional e lements can be executed simultaneously on circuits designed to solve p roblems involving more than one dimension. Experiments show that the a dditional elements required for the performance improvement have a sma ll impact on the circuit area.