CSE Ph.D. students Amol Ghoting (advisor: Srinivasan Parthasarathy) and Abhinav Vishnu (advisor: Dhabaleswar Panda) have received IBM Ph.D. Fellowships for the 2006 academic year. Both students are part of the high end systems group and these awards represent the third and fourth such awards to the group's students in the recent past. The IBM Ph.D. Fellowship program is highly competitive and prestigious for computer science and engineering graduate students, and it "honors exceptional Ph.D. students in an array of focus areas of interest to IBM and fundamental to innovation". The IBM Graduate Fellows are granted tuition, fees, and a stipend of $17,500 for one academic year, and may compete annually to be renewed for up to three years.
Amol, a member of the data mining research laboratory (DMRL) in CSE, works in the area of high performance data mining, focusing on understanding and improving the memory-system performance and caching issues related to data mining algorithms. A recent paper on this topic written together with his advisor, a fellow student, and collaborators from Intel Corporation titled "Cache-conscious Frequent Pattern Mining on a Modern Processor" was awarded the Best Paper Award at the International Conference on Very Large Data Bases 2005. Amol has accepted a Research Internship in the Data Analytics Group at IBM T.J. Watson Research Center for Summer 2006, as part of his fellowship. Abhinav, a member of the Network-Based Computing Laboratory in CSE, works in the area of high performance computing and networking, focusing on multi-rail and fault-tolerant designs for MPI on InfiniBand. The multi-rail-based designs, proposed by Abhinav and presented at the Supercomputing (SC '04) and High Performance Computing (HiPC '05) conferences, are part of the open-source MVAPICH software distribution (http://nowlab.cse.ohio-state.edu/projects/mpi-iba/) which is being used by more than 315 organizations world-wide. During Summer '05, Abhinav was invited to IBM, Austin as a summer intern to design MVAPICH support for IBM's next generation InfiniBand adapters. These designs are being incorporated into the upcoming MVAPICH releases. As a part of the IBM Fellowship, Abhinav is planning to join IBM, Poughkeepsie as an intern during Summer 2006 to design high performance MPI for next generation ultra-scale clusters (consisting of multi-thousand nodes).
This year's competition was extremely stiff, with over 500 applicants worldwide out of which roughly 50 fellows were chosen. CSE is very proud of the accomplishments of these students and we have great expectations from them in the future.