Schedule:
This two-day tutorial will run July 26-27, 12:00pm-3:25pm ET.
Abstract:
A majority of HPC system users use scripting languages such as Python to prototype their computations, coordinate their large executions, and analyze the data resulting from their computations. Python is great for these many uses, but it frequently falls short when significantly scaling up the amount of data and computation, as required to fully leverage HPC system resources. In this tutorial, we show how example computations such as heat diffusion, k-mer counting, file processing, and distributed maps can be written to efficiently leverage distributed computing resources in the Chapel, UPC++, and Fortran parallel programming models. This tutorial should be accessible to users with little-to-no parallel programming experience, and everyone is welcome. A partial differential equation problem will be shown in all three programming models along with performance and scaling results on big machines. Attendees will be shown how to compile and run these programming examples, and provided opportunities to experiment with different parameters and code alternatives while being able to ask questions and share their own observations. Come join us to learn about some productive and performant parallel programming models!
Current OLCF users with access to Frontier will be able to access a reservation on Frontier to work the examples. Current NERSC users will be able to use Perlmutter. Training accounts on Perlmutter are available for participants who do not have access to either Frontier or are not NERSC users. The examples will also be available in a Docker container and a cloud-based virtual desktop environment for access by any attendee.
Keywords:
- Basic and introductory topics for expanding broader engagement
- Software engineering for portable performance and scalability
- Parallel programming methods, models, languages and environments
- Clusters and distributed systems
The IDEAS Productivity project, in partnership with the DOE Computing Facilities of the ALCF, OLCF, and NERSC, and the DOE Exascale Computing Project (ECP), organizes the webinar series on Best Practices for HPC Software Developers.
As part of this series, we offer one-hour webinars on topics in scientific software development and high-performance computing, approximately once a month. The August webinar is titled Infrastructure for High-Fidelity Testing in HPC Facilities; and will be presented by Ryan Prout (Oak Ridge National Laboratory). The webinar will take place on Wednesday, August 9, 2023, at 1:00 pm ET.
Abstract:
The Exascale Computing Project (ECP) is investing heavily in software for exascale systems, as can be seen in the many tools, libraries and software components within ECP. In order to boost software integration across computing facilities, ECP has developed infrastructure and tools for high-fidelity testing. This infrastructure is made accessible to ECP software technology developers to provide a trusted and efficient testing environment that employs continuous integration (CI). At the core of the ECP-enabled testing infrastructure is the Jacamar CI tool. This tool allows us to link multi-tenant HPC systems to Gitlab CI workflows. This webinar will provide an overview of the ECP testing infrastructure, discuss what this could look like post-ECP, and how it could benefit other HPC facilities.
2023 HDF5 User Group Meeting (HUG23)
Registration is now open for the 2023 HDF5 User Group (HUG) Meeting being held August 16-18, 2023 at Scott Laboratory on The Ohio State University campus. This year is special, as we are celebrating 25 years of HDF5. The meeting will consist of over 25 presentations from community members, keynote speakers, lunch talks, a group dinner out, and a full day of tutorials.
Complete information, including information on the conference’s hotel room block, daily schedules, and more can be found on the conference website.
2023 HDF5 User Group Meeting (HUG23)
Registration is now open for the 2023 HDF5 User Group (HUG) Meeting being held August 16-18, 2023 at Scott Laboratory on The Ohio State University campus. This year is special, as we are celebrating 25 years of HDF5. The meeting will consist of over 25 presentations from community members, keynote speakers, lunch talks, a group dinner out, and a full day of tutorials.
Complete information, including information on the conference’s hotel room block, daily schedules, and more can be found on the conference website.
2023 HDF5 User Group Meeting (HUG23)
Registration is now open for the 2023 HDF5 User Group (HUG) Meeting being held August 16-18, 2023 at Scott Laboratory on The Ohio State University campus. This year is special, as we are celebrating 25 years of HDF5. The meeting will consist of over 25 presentations from community members, keynote speakers, lunch talks, a group dinner out, and a full day of tutorials.
Complete information, including information on the conference’s hotel room block, daily schedules, and more can be found on the conference website.
The IDEAS Productivity project, in partnership with the DOE Computing Facilities of the ALCF, OLCF, and NERSC, and the DOE Exascale Computing Project (ECP), organizes the webinar series on Best Practices for HPC Software Developers.
As part of this series, we offer one-hour webinars on topics in scientific software development and high-performance computing, approximately once a month. The September webinar is titled Simplifying Scientific Python Package Installation and Usage; and will be presented by Amiya Maji (Purdue University). The webinar will take place on Wednesday, September 13, 2023, at 1:00 pm ET.
Abstract:
With the growing popularity of Python, installation and management of Python packages in HPC environments is emerging as a critical problem for researchers; the problem is exacerbated by the need to provide consistency across traditional batch workloads and interactive notebooks. This webinar will discuss how to simplify scientific Python package installation by streamlining environment management, dependency tracking, and runtime customizations through easy-to-use tools. The webinar will discuss challenges for installing Python packages in HPC environments and present the best practices suggested by various HPC centers. Many of these best practices have been incorporated into a tool, conda-env-mod, developed by the speaker and his collaborators. HPC centers can further customize the tool and its module templates to incorporate additional software dependencies and provide descriptive help messages. The deployment of the tool has significantly reduced errors and enabled sharing of Python package installations among users. The webinar will give an overview of installing Python packages with conda-env-mod.
The IDEAS Productivity project, in partnership with the DOE Computing Facilities of the ALCF, OLCF, and NERSC, and the DOE Exascale Computing Project (ECP), organizes the webinar series on Best Practices for HPC Software Developers.
As part of this series, we offer one-hour webinars on topics in scientific software development and high-performance computing, approximately once a month. The October webinar is titled Taking HACC into the Exascale Era: New Code Capabilities, and Challenges; and will be presented by Esteban Rangel (Argonne National Laboratory). The webinar will take place on Wednesday, October 11, 2023, at 1:00 pm ET.
Abstract:
HACC (Hardware/Hybrid Accelerated Cosmology Code) is a well-established code within the US Department of Energy community, and with a long history — having run on every flagship computing system for over a decade. Often participating in early-access programs for upcoming systems, an ongoing challenge for HACC developers is to not only contend with state-of-the-art architectures, but also with their initially supported, and often novel, programming models. The increased computing power brought about by today’s exascale systems has allowed HACC to support additional baryonic physics through a newly developed Smoothed Particle Hydrodynamics (SPH) formalism called Conservative Reproducing Kernel (CRK). This webinar will discuss the challenges faced in preparing HACC for multiple exascale systems while simultaneously adding additional code capabilities, with ongoing development, all the while with a central focus on performance.
The IDEAS Productivity project, in partnership with the DOE Computing Facilities of the ALCF, OLCF, and NERSC, and the DOE Exascale Computing Project (ECP), organizes the webinar series on Best Practices for HPC Software Developers.
As part of this series, we offer one-hour webinars on topics in scientific software development and high-performance computing, approximately once a month. The November webinar is titled A cast of thousands: How the IDEAS Productivity project has advanced software productivity and sustainability; and will be presented by David Bernholdt (Oak Ridge National Laboratory). The webinar will take place on Wednesday, November 8, 2023, at 1:00 pm ET.
Abstract:
The US Department of Energy’s Exascale Computing Project (ECP) has been an unprecedented effort to establish a software ecosystem spanning 24 scientific applications, 6 co-design centers, and the supporting software technologies needed to enable leading-edge computational science and engineering research on the world’s first generation of exascale computers. ECP also presented an unprecedented challenge from the standpoint of developer productivity and the sustainability of all of that software, which led to the establishment in 2017 of the second instance of the IDEAS Productivity project, IDEAS-ECP.
Considering the scale of the ECP, involving nearly one thousand people in total, members of the IDEAS-ECP project had to think creatively about how to help so many software teams across the ECP “up their game” with respect to their software practices. This webinar will describe some of the strategies that the IDEAS team has used to pursue this goal and some of the impacts our work has had—as we are partnering with the ECP and the broader community to reduce technical risk, improve overall scientific productivity, and build a firm foundation for tackling even greater challenges in next-generation computation science.
We will wrap up with some “lessons learned” from the IDEAS experience about software stewardship and briefly consider some of the possible futures for the DOE scientific software community.
The ECP SOLLVE project, which is working to evolve OpenMP for exascale computing, invites you to participate in a new series of monthly telecons that will occur on the last Friday of every month. The next call in the series will take place on Friday, December 1st, between 11:00 am and 12:00 pm ET.
We are organizing these monthly calls so that ECP application teams may share their OpenMP experiences with the community and bring any related issues or concerns to the attention of the compiler developers and OpenMP language committee members. Application developers may treat them as office hours on all topics related to OpenMP. We expect that representatives of vendors will attend on a regular basis. Please note that attendance is open to ECP and the broader HPC community, and therefore participants should not share confidential and/or proprietary information.
Our goal is to enable application teams to be more productive using OpenMP and help make your codes portable across different vendor compilers and systems. The telecons will be conducted via Zoom. In order to receive the Zoom coordinates for the call, please fill out the following form or click “Tickets” above. Note, you will only be required to fill this form out once to receive the invite to the monthly series.
For the agenda and previous telecons’ materials please check
https://www.openmp.org/events/ecp-sollve-openmp-monthly-teleconference/
The IDEAS Productivity project, in partnership with the DOE Computing Facilities of the ALCF, OLCF, and NERSC, and the DOE Exascale Computing Project (ECP), organizes the webinar series on Best Practices for HPC Software Developers.
As part of this series, we offer one-hour webinars on topics in scientific software development and high-performance computing, approximately once a month. The November webinar is titled Secure Software Programming Practices and Development; and will be presented by Nitin Sukhija (Slippery Rock University of Pennsylvania). The webinar will take place on Wednesday, December 13, 2023, at 1:00 pm ET.
Abstract:
In scientific software development design priorities such as cost, speed, convenience, open architecture, backwards compatibility, and others often trump security requirements. Moreover, this problem is compounded by an increased dependence on integration of a myriad of software packages and technologies along with an increased need for long-term sustainable software. Thus, incorporating secure software knowledge and practices into scientific software development is becoming increasingly critical to mitigate and to defend against malicious attacks that can cause extreme damage to any piece of software, compromising integrity, authentication, and availability. The webinar aims to address this issue by providing a comprehensive overview of the secure software development process and a thorough introduction to threat modeling and security assurance testing.
