Parallel computation of matrix norm based on MapReduce

Yuwan Gu, Dongyu Zhang, Yan Chen, Bixia Chao, Yuqiang Sun

COMPUTER MODELLING & NEW TECHNOLOGIES 2015 19(2A) 13-16

School of Mathematics and Physics, ChangZhou University, Jiangsu, Changzhou213164, China

A kind of parallel programming method based on MapReduce model is proposed, in allusion to data characteristic of having specific data partitioning requirement, parallel computation of matrix norm is implemented on the platform of high-performance MapReduce. Comparing with the traditional parallel programming model, MapReduce model parallel program can satisfy to requirement of high performance numerical calculations well, its programming for simplicity and readability can improve parallel programming efficiency in effect.