Parallel Delay-and-Sum Algorithm Implemented on SuperVessel Cloud with High-Performance FPGA.

Junying Chen,Shunfeng Zhou,Huaqing Min

DOI: https://doi.org/10.1109/hpcc-smartcity-dss.2016.0144

2016-01-01

Abstract:As the rapid development of big data and cloud services, remote medical care is receiving more and more attention. In the case of remote medical ultrasound diagnoses, the real-time imaging performance is the critical implementation requirement. In order to compensate the remote network communication latency, a solution which can highly boost up the computational speed of delay-and-sum algorithm is necessary. The computational speed of the conventional computing machine cannot fulfil the high-speed computational requirement for the remote medical ultrasound diagnostic service. As a result, SuperVessel cloud with high-performance field-programmable gate arrays is adopted to implement the parallel delay-and-sum algorithm on the cloud platform. The computational speed of the high-performance SuperVessel cloud implementation is 22 times faster than the computational speed of the algorithm implementation on a central processing unit, including the data transfer delay between the central processing unit and the field-programmable gate array.