New approximate multiplier for low power digital signal processing

Abstract

In this paper a low power multiplier is proposed. The proposed multiplier utilizes Broken-Array Multiplier approximation method on the conventional modified Booth multiplier. This method reduces the total power consumption of multiplier up to 58% at the cost of a small decrease in output accuracy. The proposed multiplier is compared with other approximate multipliers in terms of power consumption and accuracy. Furthermore, to have a better evaluation of the proposed multiplier efficiency, it has been used in designing a 30-tap low-pass FIR filter and the power consumption and accuracy are compared with that of a filter with conventional booth multipliers. The simulation results show a 17.1% power reduction at the cost of only 0.4dB decrease in the output SNR.

Adaptive verifiability-driven strategy for evolutionary approximation of arithmetic circuits

“…In the comparison, we consider two approximate architectures for multipliers that are known to provide the best results, namely truncated multipliers (TMs) that ignore the values of least significant bits and broken-array multipliers (BAMs) [45]. TMs and BAMs can be parameterised to produce approximate circuits for the given bit-width and the required error.…”

A Methodology for Improvement of Roba Multiplier for Electronic Applications

“…Mandeni et al propose a bioinspired approach, where the addition results within the multiplier were approximated by the use of OR gates for the lower part of the inputs [4]. At the same time, new metrics and methodologies are proposed in [5], [6], [7] for evaluation of modeling of approximate adders. In [3], an approximate multiplier and an approximate adder based on a technique named broken-array multiplier (BAM) were proposed.…”

“…Kulkarnietal. [6] suggested an approximate multiplier factor consisting of variety a number of 2 × 2 inaccurate building blocks that saved the power facility by 31.8%-45.4% over an accurate multiplier. An approximate signed 32-bit multiplier for speculation purposes in pipelined processors was designed in [7].…”

A design methodology for approximate multipliers in convolutional neural networks: A case of MNIST

“…According to the work in [6] and [7], which investigated the recent studies in the area of approximate computing, the design of approximate multipliers has extensively been studied. Many approximate multipliers proposed calculate the upper bits exactly and lower bits approximately [8][9][10]. Yamamoto et al proposed a methodology to systematically design 8-bit approximate array multipliers by gradually removing adders, and analyzed the trade-offs among circuit area, delay, power and accuracy of the approximate multipliers [11].…”