Dask is a Python-based "flexible parallel computing library for analytic computing." Dask provides a simple interface to parallel data analysis tasks without requiring parallel computing knowledge from the user. Xarray is a Python package that provides an in-memory data model (and operations on that data) consistent with the Common Data Model used by NetCDF. Together, Dask and Xarray provide a parallel-computing platform ideal for analysis of Earth System Science data.
Speaker Description:
Dr. Kevin Paul is a scientist at the National Center for Atmospheric Research and an expert on parallel computation and workflow optimization.
Dr. Joe Hamman is a post-doctoral fellow with the National Center for Atmospheric Research and one of the lead authors of Xarray.
The use of the Python programming language has grown immensely over the past decade and has become an essential tool within education, research, and industry within the atmospheric sciences. This course aims to go beyond a basic Python introduction and help attendees advance their ability to apply Python to practical problems in meteorology. This includes topics such as remote data access, calculation of derived quantities, and plotting of these quantities on map projections.
Speaker Description:
Ryan May is a software engineer at UCAR/Unidata, working on Python software and training for the atmospheric science community. Currently, he is the core developer of the MetPy and Siphon Python packages, as well as a member of the development team for the matplotlib Python visualization library.
The current wave of advances in Deep Learning (DL) has led to many exciting challenges and opportunities for Computer Science and Artificial Intelligence researchers alike. Modern DL frameworks like Caffe/Caffe2, TensorFlow, CNTK, Torch, and several others have emerged that offer ease of use and flexibility to describe, train, and deploy various types of Deep Neural Networks (DNN) including deep convolutional nets. In this tutorial, we will provide an overview of interesting trends in DL and how cutting-edge hardware architectures are playing a key role in moving the field forward.
Speaker Description:
Ammar Ahmad Awan received his B.S. and M.S.degrees in Computer Science and Engineering from National University of Science and Technology (NUST), Pakistan and Kyung Hee University (KHU), South Korea, respectively. Currently, Ammar is working towards his Ph.D. degree in Computer Science and Engineering at The Ohio State University. His current research focus lies at the intersection of High Performance Computing (HPC) libraries and Deep Learning (DL) frameworks. He previously worked on a Java-based Message Passing Interface (MPI) and nested parallelism with OpenMP and MPI for scientific applications. He has published 14 papers in conferences and journals related to these research areas. He actively contributes to various projects like MVAPICH2-GDR (High Performance MPI for GPU clusters, OMB (OSU Micro Benchmarks), and HiDL (High Performance Deep Learning). He is the lead author of the OSU-Caffe framework (part of HiDL project) that allows efficient distributed training of Deep Neural Networks.
Dhabaleswar K. (DK) Panda is Professor and University Distinguished Scholar of Computer Science and Engineering at The Ohio State University. He leads the Network-Based Computing Research Group.
The tutorial will start with an overview of the MVAPICH2 libraries and their features. Next, we will focus on installation guidelines, runtime optimizations and tuning flexibility in-depth. An overview of configuration and debugging support in MVAPICH2 libraries will be presented. Support for GPUs and MIC enabled systems will be presented. The impact on performance of the various features and optimization techniques will be discussed in an integrated fashion. `Best Practices' for a set of common applications will be presented.
Speaker Description:
Dr. Hari Subramoni is a research scientist in the Department of Computer Science and Engineering at the Ohio State University. His current research interests include high performance interconnects and protocols, parallel computer architecture, network-based computing, exascale computing, network topology aware computing, QoS, power-aware LAN-WAN communication, fault tolerance, virtualization, big data and cloud computing. He has published over 50 papers in international journals and conferences related to these research areas. He has been actively involved in various professional activities in academic journals and conferences. He is a member of IEEE.
Dhabaleswar K. (DK) Panda is Professor and University Distinguished Scholar of Computer Science and Engineering at The Ohio State University. He leads the Network-Based Computing Research Group.
Overlapping Communication with computation Symposium
Speaker:
Julien Jaeger
With machines now gathering millions of cores, efficient data movement is becoming the challenge to feed computing units. One effective way of hiding the communication cost is to recover them with computation. On this basis, existing MPI runtimes and networks often provide a limited overlap. In this presentation, we propose to introduce our overlapping implementation, integrated in the MPC runtime. We show how this approach achieves 50%+ of fully offloaded direct buffer to buffer (zero copy) messaging on whole runs.
Speaker Description:
Julien Jaeger is Research Engineer at CEA (Commissariat a l'energie atomique et aux energies alternatives, French Alternative Energies and Atomic Energy Commission). He works on the MPC (Multi-Processor Computing) framework provides a unified parallel runtime designed to improve the scalability and performances of applications running on clusters of (very) large multiprocessor/multicore NUMA nodes.
Overlapping Communication with computation Symposium
Speaker:
Sayantan Sur
High-end computing relies on utilizing system resources effectively in order to maximize throughput. Both computation and communication make use of precious system resources, therefore it is ideal to utilize them both simultaneously. Overlapping computation with communication requires careful application programming along with appropriate support from the communication subsystem. The Open Fabrics Interface (OFI) is an effort by the Open Fabrics Alliance to provide application-centric communication interfaces. The OFI APIs provide mechanisms to overlap computation with communication.
Speaker Description:
Sayantan's specialty is in high-performance compute fabrics and the topic of performance in general. His most recent work is the definition of the Open Fabrics Interfaces (OFI). The goal of OFI is to define interfaces that enable a tight semantic map between applications and the fabric. The software interfaces have been co-designed with fabric hardware providers and application developers, with a focus on the needs of HPC users.
Sayantan's passionate about connecting large scale scientific and big data applications to the underlying system architecture by appropriate programming model design.
Find and fix your bugs easily using the highly scalable Arm DDT debugger. Learn how to prepare code for use with DDT, and how to set up remote clients to step through your code on Cheyenne using a graphical interface. Several hands-on examples will be provided to demonstrate the basic features of Arm DDT
Speaker Description:
Ryan Hulguin is an Applications Engineer for Arm. He earned a M.S. in Computational Engineering from the University of Tennessee at Chattanooga while studying CFD using model Boltzmann Equations. He then spent several years providing support for XSEDE HPC resources, before joining the Allinea software team that was eventually acquired by Arm.
Climate simulation codes, such as the Community Earth System Model (CESM), are especially complex and continually evolving. Their ongoing state of development requires frequent software verification in the form of quality assurance to both preserve the quality of the code and instill model confidence. To formalize and simplify this previously subjective and computationally expensive aspect of the verification process, we have developed a new tool for evaluating climate consistency, the CESM ensemble consistency test, referred to as CESM-ECT.
Speaker Description:
Dorit Hammerling is a researcher at the Institute for Mathematics Applied to Geosciences, in the National Center for Atmospheric Research. She is interested in spatio-temporal statistical methods applied to the geosciences, with a focus on massive data from satellites and climate models. To that end, she is interested in ways to optimally use the high performance computing facilities available at NCAR.
The talk will present an overview of the TAU Performance System to instrument, measure, and analyze the performance of parallel applications. It will show a brief demo of TAU’s ParaProf profile browser. Developers are encouraged to bring codes with them to the SEA workshop and work with the TAU developers during the week.
Speaker Description:
Sameer Shende is Director of the Performance Research Laboratory at the University of Oregon
XALT collects accurate, detailed, and continuous job-level and link-time data and stores that data in a database; all the data collection is transparent to the users. The amount of data generated can be hundreds of Gigabytes per year for a large center. The data stored can be mined to generate a picture of the compilers, libraries, and other software that users need to run their jobs successfully, highlighting the products that researchers use. We showcase how data collected by XALT can be easily mined into a digestible format by presenting data from four separate HPC centers.
Speaker Description:
Doctor Robert McLay received bachelors and masters degree from the Massachusetts Institute of Technology and his Ph.D in Engineering Mechanics from The University of Texas at Austin. His research include C++ software development, regression testing, and software tools, all related to large parallel numerical simulation codes. In particular, he has done work in parallel finite-element programs solving incompressible fluid flow and heat transfer.
His interest in software tools and support of HPC programming environments has lead to his development of Lmod, a modern replacement for Environment Modules system. Lmod's major advantages are protect all users from loading incompatible software without hindering experts. This work as lead to an interest in tracking the software usage through the module system.
