[mvapich-discuss] Final Program for MVAPICH User Group (MUG '20) Virtual Event - Free Registration

[Panda, Dhabaleswar]

The final program for the 8th annual MVAPICH User Group (MUG) meeting is included below. It will be held in a virtual manner on Monday, August 24th through Wednesday, August 26th.

Registration is free and we encourage you to register by August 21st: http://mug.mvapich.cse.ohio-state.edu/registration/.

A very special THANK YOU to our sponsors, including: ARM, Broadcom, NVIDIA, Microsoft, ParaTools, X-ScaleSolutions, Ohio Supercomputing Center and The Ohio State University.

We hope you will join us for an exciting program this year.

Thank you,
The MVAPICH Team

================================================================

                     8th Annual MVAPICH User Group (MUG) Meeting
                                            August 24-26, 2020

                                                 Final Program
                                            All times are U.S. EDT

August 24, 2020

10:00-11:00 – Tutorial: Performance Engineering with MVAPICH2 and Arm’s Scalable Vector Extension, John Linford, Arm

11:00-12:00 – Tutorial: SHARPv2: In-Network Scalable Streaming Hierarchical Aggregation and Reduction Protocol, Devendar Bureddy, NVIDIA Mellanox Network Business Unit

12:00-12:30 – Tutorial: Visualize and Analyze your Network Activities using OSU INAM (InfiniBand Networking and Monitoring tool), Hari Subramoni, The Ohio State University

12:30-1:00 – Break

1:00-2:00 - Tutorial: Best Practices for Running HPC Applications on Microsoft Azure, John Shelley and Jithin Jose, Microsoft Azure

2:00-3:00 - Tutorial: Thor: A 200G PCIe 4.0 Ethernet Controller for Mainstream HPC Markets, Hemal Shah, Karen Schramm, and Moshe Voloshin, Broadcom

3:00-4:00 - Tutorial: How to Boost the Performance of HPC/AI Applications using MVAPICH2 Library, Hari Subramoni, The Ohio State University


August 25, 2020

10:00-10:15 – Opening Remarks, Dave Hudak, Executive Director, Ohio Supercomputer Center  and Dhabaleswar K (DK) Panda, The Ohio State University

10:15-11:15 – Keynote: Deep Learning at Scale for Scientific Computing, Brian van Essen, Lawrence Livermore National Laboratory (LLNL)

11:15-12:00 - Overview of the MVAPICH Project and Future Roadmap, Dhabaleswar K (DK) Panda, The Ohio State University

12:00-12:30 - InfiniBand In-Network Computing Technology and Roadmap, Gilad Shainer, NVIDIA Mellanox Network Business Unit

12:30-1:00 – Break

1:00-1:30 - Scaling Message Passing on Amazon Web Services with Elastic Fabric Adapter, Raghunath Rajachandrasekar, AWS

1:30-2:00 - MVAPICH2 on Microsoft Azure, HPC Jithin Jose, Microsoft Azure

2:00-2:30 - Recent Advances in the HPC network stack on Arm, Pavel Shamis, ARM

2:30-3:00 - Implementing High-Performance MPI over oneAPI Level Zero, Sayantan Sur, Intel

3:00-3:30 - MVAPICH2 on Thor: High Performance MPI Meets Mainstream Ethernet
Controller, Hemal Shah, Moshe Voloshin, and Devesh Sharma, Broadcom

3:30-4:00 - Performance Engineering using MVAPICH and TAU Sameer Shende, University of Oregon and ParaTools

4:00-4:30 - Short Presentations:

Hardware Tag Matching in MVAPICH2-X, Mohammadreza Bayatpour, The Ohio State University

Efficient Support of Large Message Neighborhood Collectives in MVAPICH2-X, Mahdieh Ghazimirsaeed, The Ohio State University

Accelerating CPU-Based Training for Very large Deep Neural Networks using MVAPICH2-X, Arpan Jain, The Ohio State University


August 26, 2020

10:00-11:00 – Keynote: Computing without boundaries on Expanse, Michael L. Norman, San Diego Supercomputing Center (SDSC)

11:00-11:30 - Performance of Applications on Nurion Utilizing MVAPICH2-X, Minsik Kim, KISTI Supercomputing Center, South Korea

11:30-12:00 - Kernel-Level Support for MPI Intra-Node Communication (Post-LiMIC2): Project Overview, Hyun-Wook Jin, Konkuk University, South Korea

12:00-12:30 - Impact of SHARP and Adaptive Routing on Applications on Frontera, John Cazes, Texas Advanced Computing Center (TACC)

12:30-1:00 – Break

1:00-1:30 - Application and Microbenchmark Performance Study using MVAPICH2 and MVAPICH2-GDR on SDSC Comet and Expanse, Mahidhar Tatineni, San Diego Supercomputing Center (SDSC)

1:30-2:00 - Building Brain Circuits: Experiences with shuffling terabytes of data over MPI, Matthias Wolf, Blue Brain Project, EPFL, Switzerland

2:00-2:30 - OSU INAM at Ohio Supercomputer Center, Karen Tomko and Heechang Na, Ohio Supercomputing Center

2:30-2:45 - Cyberinfrastructure Research, Learning and Workforce Development (LWD), Alan Sussman, National Science Foundation (NSF)

2:45-3:15 - Deep Introspection for Deep Learning and Exploiting Offloading Capabilities of Bluefield Adapters: The MVAPICH2 Approach, Donglai Dai, X-ScaleSolutions

3:15-4:15 - Short Presentations

Exploiting AMD GPUs with MVAPICH2-GDR, Jahanzeb Hashmi, The Ohio State University

High-Performance Alltoall and Allgather Support for Dense GPU Systems in MVAPICH2-GDR, Kawthar Shafie Khorassani, The Ohio State University

Scaling Distributed PyTorch and Deepspeed with MVAPICH2-GDR, Quentin Anthony, The Ohio State University

Training of Very Large Pathology Images using MVAPICH2-GDR, Arpan Jain, The Ohio State University

Efficient Communication support for DASK using MVAPICH2-GDR, Aamir Shafi, The Ohio State University

Benefits of GPU-Assisted On-the-fly Message Compression with MVAPICH2-GDR, Qinghua Zhou, The Ohio State University

Parallelizing GPU-accelerated Machine Learning Applications using MVAPICH2-GDR, Aamir Shafi, The Ohio State University

4:15–4:30 Open MIC & Conclusions