Shaden Smith profile picture

I’m a PhD candidate in computer science at the University of Minnesota. I am advised by George Karypis and focus on developing parallel algorithms for large-scale tensor factorization. More broadly, I’m interested in high performance algorithms for large-scale machine learning and data analytics. Most of my research is released part of SPLATT, an open source software toolkit for sparse tensor factorization. Another recent project is FROSTT, a collection of open sparse tensor datasets.

I was at the University of Kentucky from 2009 to 2012 and earned my BSc in computer science. While there, I did research under Mirek Truszczynski and Brent Seales.

I have completed four internships, most recently at the Intel Parallel Computing Lab where I developed high performance algorithms for graph analytics and was also involved in hardware-software codesign. Before Intel, I spent the summer of 2013 at Lawrence Livermore National Laboratory (LLNL) porting LULESH to OpenACC. The two summers before LLNL (2011 and 2012), I interned at Lexmark as the developer of a particle flow modeling engine in CUDA.

Recent News