UFIT Research Computing was identified as an Intel Parallel Computing Center (IPCC), standing alongside other institutions having researchers leading their fields in modernizing software applications used in high-performance computing. There were two projects that were awarded an Intel fellowship for code modernization.
Machine learning
David Ojika
David Ojika is a postdoctoral research affiliate at the University of Florida. He specializes in using reconfigurable architectures for machine learning and heterogeneous computing. He has had several internships with Intel’s Data Center group, focusing on the Intel Xeon+FPGA platform and compute-near-memory infrastructure. He also spent a summer at Microsoft Research (Project Catapult), working on the integration of FPGAs in the cloud using Intel FPGAs. His research focuses on applied AI and end-to-end machine learning workflows for scientific computing and streaming big data analytics.
David’s team was selected as one of two finalists in the Dell EMC AI challenge and was announced as the winning team at the SC18 Supercomputing Conference in Dallas, TX.
Climate and weather
Kyueso Park, Jingyin Tang
Kyuseo Park is a second-year doctoral student with the Department of Computer & Information Science Engineering (CISE). He received his Master’s degree from New York State University in 2014, his major focus being high performance databases and spatial databases. At UF, he is working with Dr. Markus Schneider, and is involved in the research of parallel computational geometry. More specifically, his research for this project focuses on the design and implementation of multithreaded geometric algorithms on multi-core processors, intended to improve the query performance of spatial databases and geographical information systems.
Jingyin Tang is a doctoral candidate from the Department of Geography, with a concurrent Master’s degree in computer science from CISE. His research focus is on radar meteorology, tropical meteorology, and high performance computing in spatial modeling and mesoscale weather modeling. Jingyin works with Dr. Corene Matyas in the creation of spatio-temporal models of rainfall patterns of tropical cyclones making landfall in the U.S. by using WSR-88D Doppler radar observations.