Presenter

Ilya Afanasyev is a first-year PhD student of computational mathematics and computer science from Lomonosov Moscow State University. His research interests revolve around parallel computing: during his bachelor degree, he was working on acceleration of linear algebra-based industrial applications using NVIDIA GPUs. This research has been conducted in a tight partnership with the Russian company «Tesis» and NVIDIA Cuda Center of Excellence, and resulted in significant acceleration of the industrial package FlowVision. During his Master’s degree, Ilya started researching the implementation approaches to large-scale graph processing on various platforms, including newer generations of NVIDIA GPUs, Intel Xeon Phi KNL processors and NEC SX architecture. This research was based on a deep a-priory analysis of algorithms structure and architectures properties, aimed at creating efficient and high-performance implementations for each target architecture.