This paper present the existing design and review of hardware for UMHexagonS based motion estimation for H.264/AVC video compression. Four existing motion estimation architecture that implement UMHexagonS are presented, analysed and compared to show the area to tackle for future improvements. The presented architectures are also compared against our proposed architecture. The results are compared in term of gate count and throughput with emphasize on throughput as it shows the speed of the architecture. The proposed architecture gives a promising result compared with some of the reviewed architecture. It shows an improvement of 91% compared to the full search architecture and 20% compared to a basic UMHexagonS architecture.