PAW-ATM 2023: Parallel Applications Workshop, Alternatives To MPI+X

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.

Program

9:00 – 9:02 Karla V. Morris Wright, Elliott Slaughter, Engin Kayraklioglu, Irene Moulitsas, Bill Long, and Daniele Lezzi
“PAW-ATM2023 Introduction”

9:02 – 10:00 Survey of Technologies for Developers of Parallel Applications:
Open discussion after presentations on SHMEM, Julia, Swift/T, and cuNumeric
Session Chair: Jan Ciesko – Sandia National Laboratories
Barbara Chapman – Hewlett Packard Enterprise
Justin Wozniak – Argonne National Laboratory
Johannes Blaschke – Lawrence Berkeley National Laboratory
Rohan Yadav – NVIDIA

10:00 – 10:30 Morning Break (30 min)

10:30 – 12:30 Session 1
Session Chair: Henry Monge Camacho – Oak Ridge National Laboratory
10:30 – 10:54 Tom Westerhout and Bradford L. Chamberlain
“Implementing Scalable Matrix-Vector Products for the Exact Diagonalization Methods in Quantum Many-Body Physics”

10:54 – 11:18 Scott Bachman, Rebecca Green, Anna Bakker, Helen Fox, Sam Purkis, and Ben Harshbarger
“High-Performance Programming and Execution of a Coral Biodiversity Mapping Algorithm Using Chapel”

11:18 – 11:42 Jiakun Yan, Hartmut Kaiser, and Marc Snir
“Design and Analysis of the Network Software Stack of an Asynchronous Many-Task System – The LCI Parcelport of HPX”

11:42 – 12:06 Marcin Rogowski, Jeff R. Hammond, David E. Keyes, and Lisandro Dalcin
“shmem4py: High-Performance One-Sided Communication for Python Applications”

12:06 – 12:30 James Psota, and Armando Solar-Lezama
“Pure: Evolving Message Passing to Better Leverage Shared Memory Within Nodes”

12:30 – 2:00 Lunch Break (90 min)

2:00 – 3:00 Session 2
Session Chair: Damian Rouson – Lawrence Berkeley National Laboratory
Distinguished Speaker: Ethan Gutmann – National Center for Atmospheric Research
“Trials, Tribulations and Joys of Developing with Alternative Parallel Frameworks”

3:00 – 3:30 Afternoon Break (30 min)

3:30 – 3:54 Session 3
Session Chair: Engin Kayraklioglu
Julian Bellavita, Mathias Jacquelin, Esmond Ng, Dan Bonachea, Johnny Corbino, and Paul H Hargrove
“symPACK: A GPU-Capable Fan-Out Sparse Cholesky Solver”

3:54 – 5:30 Panel Discussion: Charting a Path to Success with Alternatives to MPI+X
Panel Chair: Michelle Strout – Hewlett Packard Enterprise
Panelists:
Ewa Deelman – University of Southern California
Jonathan Graham – Los Alamos National Laboratory
Ethan Gutmann – National Center for Atmospheric Research
Laxmikant Kale – University of Illinois Urbana-Champaign
George Karypis – University of Minnesota