Program
Program
Conference proceedings available with open access at: https://dl.acm.org/doi/proceedings/10.1145/3555819
Monday, September 26, 2022
9:00 AM - 9:10 AM Welcome
9:10 AM - 10:00 AM Keynote: 30 Years of MPI and many more to come… - Rajeev Thakur
10:00 AM - 10:30 AM Morning Break
10:30 AM - 12:00 PM Session I - Collectives (Session Chair: Sidharth Kumar)
10:30 AM - 11:00 AM A Locality-Aware Bruck Allgather - Amanda Bienz, Shreeman Gautam and Amun Kharel
11:00 AM - 11:30 AM Efficient Process Arrival Pattern Aware Collective Communication for Deep Learning - Pedram Mohammadalizadehbakhtevari, Amirhossein Sojoodi, Yiltan Hassan Temucin and Ahmad Afsahi
11:30 AM - 12:00 PM Applying on Node Aggregation Methods to MPI Alltoall Collectives: Matrix Block Aggregation Algorithm - George Chochia, David Solt and Joshua Hursey
12:00 PM - 1:30 PM Lunch
1:30 PM - 3:00 PM Session II - Resource Management and Hybrid Programming (Session Chair: Amanda Bienz)
1:30 PM - 2:00 PM MPIX Stream: An Explicit Solution to Hybrid MPI+X Programming - Hui Zhou, Ken Raffenetti, Yanfei Guo and Rajeev Thakur
2:00 PM - 2:30 PM Towards Dynamic Resource Management with MPI Sessions and PMIx - Dominik Huber, Isaias Compres, Maximilian Streubel, Martin Schulz, Martin Schreiber and Howard Pritchard
2:30 PM - 3:00 PM Invited Talk: MANA/CRAC: A Preprocessor for Safe Transparent Checkpointing of MPI and CUDA - Gene Cooperman and Twinkle Jain
3:00 PM - 3:30 PM Afternoon Break
3:30 PM - 5:00 PM Session III - Correctness and Error Handling (Session Chair: Wesley Bland)
3:30 PM - 4:00 PM Towards a Hybrid MPI Correctness Benchmark Suite - Tim Jammer, Alexander Hück, Jan-Patrick Lehr, Joachim Protze, Simon Schwitanski and Christian Bischof
4:00 PM - 4:30 PM Error-Handling in MPI Continuations - Joseph Schuchart, Aurelien Bouteiller and George Bosilca
4:30 PM - 5:00 PM Invited Talk: Evolving the Message Passing Interface for Future Architectures: MPI 4.0 and beyond - Martin Schulz
Tuesday, September 27, 2022
9:00 AM - 9:10 AM Welcome
9:10 AM - 10:00 AM Keynote by Ron Brightwell
10:00 AM - 10:30 AM Morning Break
10:30 AM - 12:00 PM Session IV -Applications (Session Chair: Sayan Ghosh)
10:30 AM - 11:00 AM Enabling Global Addressing in MPI Applications - Jean-Baptiste Besnard, Sameer Shende, Allen D. Malony, Julien Jaeger and Marc Perache
11:00 AM - 11:30 AM Distributed Acceleration of Adhesive Dynamics Simulations - Daniel F. Puleri, Aristotle X. Martin and Amanda Randles
11:30 AM - 12:00 PM Invited Talk By Patrick Bridges
12:00 PM - 1:30 PM Lunch
1:30 PM - 3:00 PM Session V - Panel - Challenges and opportunities in supporting heterogeneous architectures and programming paradigms for Exascale Computing.
Moderator: Ron Brightwell
Panelists: Howard Pritchard Jr., Anthony Skjellum, Rajeev Thakur
3:00 PM - 3:30 PM Afternoon Break
3:30 PM - 5:00 PM Session VI - Posters
3:30 PM - 4:00 PM Poster Lightning Talks
Functional Implementation of Partitioned Collective Communication Primitives
Muzakhir Amanzholov, Derek Schafer, Anthony Skjellum, Sheikh Ghafoor and Purushotham Bangalore
A System-Wide Communication to Couple Multiple MPI Programs for Heterogeneous Computing
Shinji Sumimoto, Takashi Arakawa, Yoshio Sakaguchi, Hiroya Matsuba, Hisashi Yashiro, Toshihiro Hanawa and Kengo Nakajima
Kokkos-Enhanced ExaMPI
Evan Suggs, Derek Schafer and Anthony Skjellum
4:00 PM - 5:00 PM Posters
Wednesday, September 28, 2022
9:00 AM - 10:00 AM Tutorials: Effective Multithreaded Communication for Next Generation Scientific Applications by Matthew G. F. Dosanjh, W. Pepper Marts, Jan Ciesko
10:00 AM - 10:30 AM Morning Break
10:30 AM - 12:00 PM Tutorials continue
12:00 PM - 1:30 PM Lunch
2:00 PM - 3:00 PM EuroMPI/USA and IWOMP Joint Keynote Talk
OpenMP and MPI at Exascale - Early Experiences on Frontier - Al Geist