DESIGN OF AN EFFICIENT LOW ENERGY SIGNAL AWARE APPROXIMATE MULTIPLIER

ECE

ABSTRACT This research paper represents the Error-tolerant applications like multimedia and machine learning frequently use estimated computing, that relaxes full correctness to obtain better performance or decreased power usage. For major of these applications, multiplication is an essential operation. The 4:2 compressor used in this study fairly distributes errors in both positive and negative sign. Two 8 x 8 impartial approximate multipliers (UBAMs), based on the unbiased 4:2 compressor, are created to satisfy different accuracy (or power) needs. According to experimental data, one of the suggested designs offers a 39 percent smaller PDP (power-delay product) and roughly a 46 percent smaller energy-delay product when compared to earlier approximation multipliers (EDP), while the additional offers a decreased NMED (normalized mean error distance). The recommended multipliers outperform prior approximations in image filtering applications because they produce outputs of higher quality while using less energy.

This work is licensed under a Creative Commons Attribution-ShareAlike 4.0 International License.