This model is likely the best in terms of cochlea-like performance at the present time. We would choose to use this as our standard front-end, but it is computationally difficult compared to the nonlinear gamma-tone filter. Therefore, we plan to build the TWAMP using several circuit forms. We already have a 20-section discrete component model that works reasonably well over a narrow frequency band. Even bigger than the gamma-tone filter on a per-section basis in discrete component form, simulations show that many more sections are needed (e.g. > 256) to represent well the entire cochlea. The initial phase of the hardware implementations will explore the use of DSP chips. Next, we shall explore the switched capacitor approach and other suprathreshold technologies. One of particular technological interest uses an integrated CMOS inductor (patented by Hubbard) that has not been used previously except in small test designs. The TWAMP is potentially of high commercial value for its filter characteristics. A full-scale design example of the integrated inductor could well open the door for CMOS analog designs using inductors of sufficient size to extend the frequency limits to within the audio range. Currently, inductors are used principally in the GHz frequency range.