APPROACH TO ACCELERATE ALGORITHMS TO SOLVE LOGISTIC PROBLEMS WITH GPGPU

Abstract

Nowadays algorithms that were used to solve logistic problems are more and more difficult. Thanks to the globalization the networked multiple supply chains becomes even more complex. Moreover the data available from these networks growing to unprecedented levels. The fast decision making is business critical. The article is about the approach of use the GPU to accelerate algorithms to solve logistic problems faster. After some possible areas of this field are discussed we tried out the popular scientific GPU programming library named Torch. Finally test results of the measurements are presented.