An algorithm for generation of behavioral shape functions

In this paper, we present a new algorithm for analyzing performance/cost tradeoffs in interactive synthesis of DSP algorithms. Our algorithm handles both memories with different access times and pipelined units with different numbers of stages. The output is a shape function illustrating the performance vs. cost tradeoff. We have tested this algorithm on several benchmarks including an FIR filter and a robot kinematics example. Results show that the average difference in cost, as compared to manual designs, is 0.41%, while the average performance difference is 4.90% without memory access times and 0.77% with memory access times.<
>