3D-Sorter: 3D Design of a Resource-Aware Hardware Sorter for Edge Computing Platforms Under Area and Energy Consumption Constraints

Abstract

In this paper, we proposed a 3-dimensional hardware sorting architecture (3D-Sorter), based on MultiDimensional Sorting Algorithm (MDSA). the proposed architecture transforms a sequence of input records into a 3-dimensional matrix. Records of every dimension are sorted in several MDSA phases, using partial sorting methods. Our synthesis results, provided by Xilinx Vivado indicate that the 3D-Sorter design decreases the number of Look-Up Tables (LUT) and registers by 54% and 42.7%, compared to the state-of-the-art hardware sorter. Also, the power consumption is reduced by 48.15% on average. The results show that the proposed architecture is a remarkable power/area saving for edge components.