High-performance approximate multiplier based on operand cutting and calculation method thereof

Abstract

The invention discloses a high-performance approximate multiplier based on operand cutting and a calculation method of the high-performance approximate multiplier. The approximate multiplier comprises a partial product generation module, a partial product tree compression module, a carry adder module and a shifting module. The partial product generation module comprises an operand cutting module and a bit product module, the operand cutting module performs cutting operation on a multiplier and a multiplicand to obtain an approximate operand, and the bit product module obtains a corresponding partial product; the partial product tree compression module is used for obtaining two final summed addend numbers; the carry adder module is used for finally adding two adders, and the shifting module is used for shifting the output of the carry adder module to obtain a solved binary result. According to the invention, the area, the delay and the power consumption of the multi-bit multiplier can be greatly reduced, and meanwhile, better accuracy can be kept.

Description

technical field [0001] The invention belongs to the technical field of integrated circuits, in particular to a high-performance multiplier based on operand trimming and a calculation method. Background technique [0002] In recent years, with the rapid development of big data, the image information that needs to be processed increases exponentially. Basic image processing includes image filtering, image edge detection and sharpening. The image filtering operation needs to consume a large number of multiplying and adding operation units, and the precision of the result is not very strict in the filtering process. If an accurate multiplier is used, the calculation unit will be too large, and the multiplier will be flipped frequently, making the power consumption problem of the image processing unit prominent. At the same time, since most of the edge computing terminals are mobile devices, their power supply is limited, and the problem of device power consumption is particular...

AI Technical Summary

Problems solved by technology

The image filtering operation needs to consume a large number of multiplication and addition operation units, and the accuracy requirements for the results in the filtering process are not very strict

If the precise multiplier is used, the calculation unit will be too large, and the multiplier flips frequently, which makes the power consumption of the image processing unit prominent.

At the same time, because most of the terminals of edge computing are mobile devices, their power supply is limited, and the power consumption of the devices is particularly prominent.

Images