“IPDPS is one of the premier parallel and distributed computing conferences in the world that provides broad coverage of all areas in high performance computing (HPC) and parallel computing,” said CSE Professor Ümit V. Çatalyürek, one of the leaders of Georgia Tech’s participation at this year’s symposium.

“With seven papers out of 103 in the main conference alone, and many other invited talks, plenary panels and committee services, CSE researchers not only continue to push the research frontier in HPC and parallel processing, but also demonstrate their commitment to serve the parallel processing community at large,” he said.

Several Georgia Tech faculty, including Çatalyürek, are serving on IPDPS committees and workshops.

• Srinivas Aluru – HiCOMB Workshop Steering Committee
• David A. Bader– IPDPS Steering Committee, HiCOMB Workshop Steering Committee
• Ümit V. Çatalyürek - Technical Committee on Parallel Processing (TCPP) Chair,IPDPSSteering Committee, Parallel and Distributed Scientific and Engineering Computing Panelist, Algorithms Primary Committee
• Ada Gavrilovska– Regular Program Committee, HPBDC Workshop Panel Moderator
• Vivek Sarkar– Regular Program Committee
• Richard Vuduc – iWAPT Program Committee, iWAPT Steering Committee

IPDPS also serves as the flagship activity of the TC on Parallel Processing (TCPP) which presents another addition to the Georgia Tech at IPDPS roster. CSE Professor and Institute for Data Engineering and Science Co-executive Director Srinivas Aluru was named as this year’s IEEE TCPP Outstanding Service Award winner in recognition of his professional service roles that have had a major impact on the parallel processing research community at large. 

Georgia Tech’s presence at IPDPS this year includes 10 researchers from CSE and three researchers from the School of Computer Science (SCS).

Georgia Tech’s research:

• A scalable clustering-based task scheduler for homogeneous processors using DAG partitioning 
• Accelerating Sequence Alignment to Graphs
• Asynchronous Multigrid Methods 
• Efficient Architecture-Aware Acceleration of BWA-MEM for Multicore Systems 
• Load-Balanced Sparse MTTKRP on GPUs
• Overlapping Communications with Other Communications and its Application to Distributed Dense Matrix Computations 
• ParILUT -  A Parallel Threshold ILU for GPUs