Program

09:00 am - Introduction
09:00 am - 10:00 am - Keynote - Designing High-Performance and Scalable Middleware for HPC and AI: Challenges and Opportunities by DK Panda

10:00 am - 10:30 am - Morning Break

Session #1
10:30 am - 11:00 am - Paper #1 - Leveraging Interconnect QoS Capabilities for Congestion-Aware MPI Communication by Khalilov, Slinka, Zhang
11:00 am - 11:30 am - Paper #2 - Partitioned Collective Communication by Holmes, Skjellum, Jaeger, Grant, Bangalore, Dosanjh, Bienz, Schafer
11:30 am - 12:00 pm - Paper #3 - Overalapping Communication and Computation with ExaMPI's Strong Progress and Modern C++ Design by Schafer, Hines, Suggs, Rüfenacht, Skjellum

12:00 pm to 12:30 pm - ExaMPI Community Discussion

12:30 pm to 02:00 pm - Lunch

Session #2
02:00 pm - 02:30 pm - Paper #4 - Toward Modern C++ Language Support for MPI by Ghosh, Alsobrooks, Rüfenacht, Skjellum, Bangalore, Lumsdaine
02:30 pm - 03:00 pm - Paper #5 - A FACT-Based Approach: Making ML Collective Autotuning Feasible on Exascale Systems by Wilkins, Guo, Thakur, Hardavellas, Dinda, Si

03:00 pm - 03:30 pm - Afternoon Break

Session #3
03:30 pm - 04:00 pm - Paper #6 - Accelerating Multi-Process Communication for Parallel 3-D FFT by Ayala, Tomov, Stoyanov, Haidar, Dongarra
04:00 pm - 04:30 pm - Short Paper #1 - A Benchmark to Understand Communication Performance in Hybrid MPI and GPU Applications by Haskins, Bridges, Levy, Ferreira