On July 2014 Volodymyr Turchenko, Associate Professor at Information Computing Systems and Control, took part in the annual International Conference on High Performance Computing and Simulation “The 2014 International Conference on High Performance Computing & Simulation HPCS 2014”, held 21-25 July 2014 in Bologna, Italy (http: //hpcs2014.cisedu.info/).
One of the special sessions – “Parallel computing for neural networks” was organized by As. Prof. Turchenko as part of its ongoing research on preparing a doctoral thesis. The problem of speeding up the learning process of artificial neural networks (great time learning – this is one of their main drawbacks) that are able to perform intelligent tasks such as pattern recognition, prediction, optimization, management, etc., is particularly relevant today. Now there are new high-performance hardware, including multi (manycore) processors and graphics processor unit, but new technologies and algorithms for parallelization are not yet sufficiently investigated for such neural network algorithms.