DK Panda from Ohio State University gave this talk at the Stanford HPC Conference. “This talk will focus on a range of solutions being carried out in my group to address these challenges. The solutions will include: 1) MPI-driven Deep Learning, 2) Co-designing Deep Learning Stacks with High-Performance MPI, 3) Out-of- core DNN training, and 4) Hybrid (Data and Model) parallelism. Case studies to accelerate DNN training with popular frameworks like TensorFlow, PyTorch, MXNet and Caffe on modern HPC systems will be presented.”
How to Achieve High-Performance, Scalable and Distributed DNN Training on Modern HPC Systems
May 3, 2020 by
Designing Scalable HPC, Deep Learning, Big Data, and Cloud Middleware for Exascale Systems
October 18, 2019 by
DK Panda from Ohio State University gave this talk at the UK HPC Conference. “This talk will focus on challenges in designing HPC, Deep Learning, Big Data and HPC Cloud middleware for Exascale systems with millions of processors and accelerators. For the HPC domain, we will discuss about the challenges in designing runtime environments for MPI+X (PGAS – OpenSHMEM/UPC/CAF/UPC++, OpenMP, and CUDA) programming models by taking into account support for multi-core systems (Xeon, ARM and OpenPower), high-performance networks, and GPGPUs (including GPUDirect RDMA).”
Checkpointing the Un-checkpointable: MANA and the Split-Process Approach
September 8, 2019 by
Gene Cooperman from Northeastern University gave this talk at the MVAPICH User Group. “This talk presents an efficient, new software architecture: split processes. The “MANA for MPI” software demonstrates this split-process architecture. The MPI application code resides in “upper-half memory”, and the MPI/network libraries reside in “lower-half memory”.
The ABCI Supercomputer: World’s First Open AI Computing Infrastructure
September 4, 2019 by
Shinichiro Takizawa from AIST gave this talk at the MVAPICH User Group. “ABCI is the world’s first large-scale Open AI Computing Infrastructure, constructed and operated by AIST, Japan. It delivers 19.9 petaflops of HPL performance and world’ fastest training time of 1.17 minutes in ResNet-50 training on ImageNet datasets as of July 2019. In this talk, we focus on ABCI’s network architecture and communication libraries available on ABCI and shows their performance and recent research achievements.”
Offering Bare-Metal Performance and Scalability on Cloud: The Azure-HPC Approach
September 2, 2019 by
Jithin Jose from Microsoft gave this talk at the MVAPICH User Group. “This talk focuses on how HPC offerings in Azure address these challenges and explains the design pillars that allow Microsoft to offer “bare-metal performance and scalability” on the Microsoft Azure Cloud. This talk also covers the features of latest Microsoft Azure HPC offerings and provides in-depth performance insights and recommendations for using MVAPICH2 and MVAPICH2-X on Microsoft Azure.”
Video: Managing HPC Software Complexity with Spack
September 1, 2019 by
Greg Becker from LLNL gave this talk at the MVAPICH User Group. “Spack is an open-source package manager for HPC. This presentation will give an overview of Spack, including recent developments and a number of items on the near-term roadmap. We will focus on Spack features relevant to the MVAPICH community; these include Spack’s virtual package abstraction, which is used for API-compatible libraries including MPI implementations, package level compiler wrappers, and packages which modify other package’s build environments.”
Overview of the MVAPICH Project and Future Roadmap
August 27, 2019 by
DK Panda gave this talk at the MVAPICH User Group. “This talk will provide an overview of the MVAPICH project (past, present, and future). Future roadmap and features for upcoming releases of the MVAPICH2 software family (including MVAPICH2-X and MVAPICH2-GDR) for HPC and Deep Learning will be presented. Features and releases for Microsoft Azure and Amazon AWS will also be presented. Current status and future plans for OSU INAM, OMB, and Best Practices Page will also be presented.”
Agenda Posted: MVAPICH User Group (MUG) Meeting in Ohio
July 28, 2019 by
The MVAPICH User Group Meeting (MUG 2019) has published its Speaker Agenda. The event will take place from August 19-21 in Columbus, Ohio. “MUG aims to bring together MVAPICH2 users, researchers, developers, and system administrators to share their experience and knowledge and learn from each other. The event includes Keynote Talks, Invited Tutorials, Invited Talks, Contributed Presentations, Open MIC session, hands-on sessions MVAPICH developers, etc.”
How to Design Scalable HPC, Deep Learning and Cloud Middleware for Exascale Systems
March 4, 2019 by
DK Panda from Ohio State University gave this talk at the Stanford HPC Conference. “This talk will focus on challenges in designing HPC, Deep Learning, and HPC Cloud middleware for Exascale systems with millions of processors and accelerators. For the HPC domain, we will discuss about the challenges in designing runtime environments for MPI+X (PGAS – OpenSHMEM/UPC/CAF/UPC++, OpenMP, and CUDA) programming models taking into account support for multi-core systems (Xeon, OpenPower, and ARM), high-performance networks, GPGPUs (including GPUDirect RDMA), and energy-awareness.”
OSC Hosts MVAPICH Users Group this week
August 7, 2018 by
A broad array of HPC enthusiasts have gathered at the Ohio Supercomputer Center this week for the sixth meeting of the MVAPICH Users Group (MUG). “The Network-Based Computing Research Group is lead by DK Panda, a professor and university distinguished scholar of computer science at The Ohio State University, created and enhances the popular MVAPICH HPC system software package.”
