• Sony Merugumalla
• Dr .B. Sarala

Deep learning has proved to be a highly efficient technology because of its capability to process large amounts of data. Traditional methods are no longer as popular, especially when it comes to pattern recognition. Convolution Neural Networks are a type of deep neural network that are widely used. Designing a reconfigurable systolic array is used to modify the deep learning architecture's precision level (input-output bit width) and support for both signed and unsigned image convolution operations. Improved distributed arithmetic (DA) models are used to CNN that can speed up computation and configuration levels. Employing DA logic to replace all multiplier units with shifters and adders which reduces implementation complexity.

Convolution neural network, Systolic array, Processing Elements

"Implementation of systolic array architecture using distributive arithmetic logic for convolution neural networks (CNN)", International Journal of Emerging Technologies and Innovative Research (www.jetir.org), ISSN:2349-5162, Vol.9, Issue 12, page no.d92-d100, December-2022, Available :http://www.jetir.org/papers/JETIR2212316.pdf

"Implementation of systolic array architecture using distributive arithmetic logic for convolution neural networks (CNN)", International Journal of Emerging Technologies and Innovative Research (www.jetir.org | UGC and issn Approved), ISSN:2349-5162, Vol.9, Issue 12, page no. ppd92-d100, December-2022, Available at : http://www.jetir.org/papers/JETIR2212316.pdf