In imprecision-tolerant applications such as image processing and machine learning, imprecision can be tolerated because of the nature of the application itself or the limitation of human senses. By using the approximate computation in these applications, significant po
More
In imprecision-tolerant applications such as image processing and machine learning, imprecision can be tolerated because of the nature of the application itself or the limitation of human senses. By using the approximate computation in these applications, significant power, delay, or area reductions can be achieved. In this paper, two approximate full adders and an approximate adder, with low error are proposed. The effects of die-to-die (D2D) process variation on the threshold voltage of approximate circuits have been evaluated. For evaluating the accuracy and the variability, these approximate full adders have been used and analyzed in the ripple carry adder structure, image Sharpening and image Smoothing algorithms. In terms of power-delay-product (PDP), accuracy, and area for uniformly distributed inputs, the proposed approximate full adder 1, exhibits the best performance, and the proposed approximate full adder 2 and the proposed approximate adder, show the best peak-signal-to-noise ratio (PSNR) for real images.
Manuscript profile
