Loading…
Back To Schedule
Tuesday, September 15 • 3:30pm - 4:00pm
High-Performance MPI and Deep Learning with Introspection on OpenPOWER Platform - Dhabaleswar K (DK) Panda, X-ScaleSolutions and The Ohio State University & Donglai Dai, X-ScaleSolutions

Sign up or log in to save this to your schedule, view media, leave feedback and see who's attending!

Feedback form is now closed.

This talk will focus on high-performance and scalable middleware for Message Passing Interface (MPI) and Deep Learning on OpenPOWER platform with NVIDIA GPGPUs and RDMA-enabled interconnects (InfiniBand and RoCE). The focus will be on two packages with commercial support being available from X-ScaleSolutions. The first package will focus on the OSU MVAPICH2 MPI libraries and their capabilities for high-performance computing with both CPUs (OpenPOWER) and GPUs (NVIDIA). The second package will focus on tight integration between the OSU MVAPICH2-GDR MPI library and the Horovod stack to provide high-performance and scalable Deep Learning (DL) with deep introspection (DI) capabilities for DL frameworks like TensorFlow, PyTorch and MXNet. The DI capabilities allow DL users and runtime developers to easily optimize their DL applications on modern systems. Performance results from the ORNL SUMMIT system (#2nd) and Lassen (#14th) with thousands of GPUs and POWER9 CPUs will be presented.

Continue the conversation in Slack

Speakers
DK

Dhabaleswar K (DK) Panda

Professor and University Distinguished Scholar; Founder and CEO (X-ScaleSolutions), X-ScaleSolutions and The Ohio State University
DD

Donglai Dai

Chief Engineer, X-ScaleSolutions
Dr. Donglai Dai is a Chief Engineer at X-ScaleSolutions and leads company’s R&D team. His current work focuses on developing scalable efficient communication libraries and performance analysis tools for distributed and parallel HPC and deep learning applications on HPC systems... Read More →



Tuesday September 15, 2020 3:30pm - 4:00pm CDT
Track 4
  AI  Software
  • See Session Slides yes