2021 International Workshop on Performance, Portability & Productivity in HPC

Program

The P3HPC workshop will be an all-day event on Sunday, 14 November 2021 from 9:00AM - 5:30PM Central.

The full proceedings are available here.

If prompted for a username and password, please use scwpub21 and conf//.

09:00 - 10:00
C++ Standard Parallelism
Bryce Adelstein Lelbach


Break

10:00 - 10:30


Session 1

10:30 - 11:00
Revisiting a Metric for Performance Portability
John Pennycook, Intel Corporation


11:00 - 11:30
Mamba: Portable Array-based Abstractions for Heterogeneous High-Performance Systems
Tim Dykes, Hewlett Packard Enterprise


11:30 - 12:00
oneAPI Open-Source Math Library Interface
Mariia Krainiuk, Intel Corporation


12:00 - 12:30
Analyzing Reduction Abstractions Capabilities
Tom Deakin, University of Bristol


Break

12:30 - 14:00


Session 2

14:00 - 14:30
Evaluation of Performance Portable Programming Models and Framework across AMD, Intel and NVIDIA GPUs using Applications and Mini-Apps
JaeHyuk Kwack, Argonne National Laboratory


14:30 - 15:00
Case Study of Using Kokkos and SYCL as Performance-Portable Frameworks for MILC-DSLASH Benchmark on NVIDIA, AMD and Intel GPUs
Amanda Dufek, Lawrence Berkeley National Laboratory


Break

15:00 - 15:30


Session 3

15:30 - 15:55
Evaluating Performance and Portability of a Core Bioinformatics Kernel on Multiple Vendor GPUs
Muhammad Haseeb, Florida International University


15:55 - 16:20
Optimization Strategy for a Performance Portable Vlasov Code
Yuuichi Asahi, Japan Atomic Energy Agency


Panel

16:20 - 17:30
Healing the Gaps and Growing the Scope of Performance, Portability and Productivity
C.J. Newburn, Adam Kunen, Bryce Adelstein Lelbach, Christian Trott, Tom Deakin, Bo Zhang