Parallel Applications Workshop, Alternatives To MPI+X at SC24
Dan Bonachea
Alternatives to MPI+X are worth exploring as programmer productivity becomes a major component of the time to science. Alternatives include parallel programming languages (e.g. Chapel, Regent, Fortran 2018), general purpose libraries (e.g. Charm++, COMPSs, HPX, Legion, UPC++), and domain specific libraries (e.g. Arkouda, Dask, Spark). With many options to choose from, it is hard for programmers to know which alternative models are appropriate for their application and for programming model developers to understand the opportunities for improvement.
Through discussion of specific applications, PAW-ATM brings together application experts and programming model developers to improve applications and models.
WORKSHOP PROGRAM
9:00 - 9:02 “PAW-ATM2024 Introduction” Karla V. Morris Wright, Elliott Slaughter,
Engin Kayraklioglu, Irene Moulitsas, Bill Long, Daniele Lezzi and Kenjiro Taura
Session Chair: Brad Richardson – Lawrence Berkeley National Laboratory
9:02 - 9:45 “Survey of Technologies for Developers of Parallel Applications”
Wonchan Lee – NVIDIA
Damian Rouson – Lawrence Berkeley National Laboratory
9:45 - 10:00 User Experience: “Resource Adaptivity at Task-Level” Jonas Posner
10:00 - 10:30 Morning Break (30 min)
Session Chair: Quincey Koziol – NVIDIA
10:30 - 10:50 “Speeding-Up LULESH on HPX: Useful Tricks and Lessons Learned using a Many-Task-Based Approach” Torben Kalkhof, and Andreas Koch
10:50 - 11:10 “Lamellar: A Rust-based Asynchronous Tasking and PGAS Runtime for High Performance Computing” Ryan D. Friese, Roberto Gioiosa, Joseph Cottam, Erdal Mutlu, Gregory Henselman-Petrusek, Polykarpos Thomadakis, and Mark Raugas
11:10 - 11:30 “Applying a Task-Based Approach to Distributed Machine Learning Workflows” Fernando Vazquez-Novoa, Daniele Lezzi, Francesc Lordan, Fatemeh Baghdadi, and Davide Cirillo
11:30 - 11:50 “Accelerating Multi-GPU Embedding Retrieval with PGAS style Communication for Deep Learning Recommendation Systems” Yuxin Chen, Aydin Buluc, Katherine Yelick, and John Owens
11:50 - 12:10 “Mitigating synchronization bottlenecks in high-performance actor-model-based software” Kyle Klenk, Mohammad Mahdi Moayeri, Junwei Guo, Martyn P. Clark, and Raymond J. Spiteri
12:10 - 12:30 “Intel SHMEM: GPU-initiated OpenSHMEM using SYCL” Alex Brooks, Philip Marshall, David Ozog, Md. Wasi-ur- Rahman, Lawrence Stewart, and Rithwik Tom
12:30 - 2:00 Lunch Break (90 min)
Session Chair: Michelle Strout – Hewlett Packard Enterprise
2:00 - 2:45 Distinguished Speaker: Eric Laurendeau – Polytechnique Montreal
“A case study for using Chapel within the global aerospace industry”
2:45 - 3:00 User Experience: “Exploring Suffix Array Algorithms in Chapel” Michael P. Ferguson, Bonnie Hurwitz, and Shreyas Khandekar
3:00 - 3:30 Afternoon Break (30 min)
3:30 - 3:45 Session Chair: Alex Aiken – Stanford University
3:30 - 3:45 User Experience: “Just Write Fortran: Experiences with a Language-Based Alternative to MPI+X” Baboucarr Dibba, Katherine Rasmussen, Brad Richardson, Damian Rouson, David Torres, Yunhao Zhang, Ethan Gutmann, Kareem Ergawy and Michael Klemm
3:45 - 5:30 Panel Discussion: “Alternative programming models for applications at scale”
Panel Chair: Christine Sweeney – Los Alamos National Laboratory
Panelists:
Jan Ciesko – Sandia National Laboratories
Nils Deppe – Cornell University
Jason DeVinney – Center for Computing Sciences
Eric Laurendeau – Polytechnique Montreal
Julian Samaroo – Massachusetts Institute of Technology
Best Regards,
PAW-ATM2024 Organizing Committee