Exascale Computing Project Podcast: Recent Episodes
Exascale Computing Project
The Exascale Computing Project (ECP) is accelerating delivery of a capable exascale computing ecosystem to provide breakthrough solutions that will address America's most critical challenges in scientific discovery, energy assurance, economic competitiveness, and national security. Let’s Talk Exascale explores Application Development, Software Technology, and Hardware and Integration—focus areas of the ECP.
Episode notes: Jack Dongarra says ECP has been a great success in terms of human and technical accomplishments but post-project follow-on is critical.
Episode notes: Lawrence Livermore National Lab is preparing for El Capitan, the National Nuclear Security Administration’s first exascale supercomputer.
Episode notes: The CANcer Distributed Learning Environment (CANDLE) project provides deep-learning computing methodologies for accelerating cancer research.
Episode notes: Katherine Riley leads a team of computational science experts who work with facility users to maximize their use of ALCF computing resources.
Episode notes: Richard Gerber ensures the National Energy Research Scientific Computing Center remains responsive to the needs of scientific researchers.
Episode notes: Bronson Messer, director of science for the Oak Ridge Leadership Computing Facility, has an expert perspective on supercomputing for science.
Episode notes: ECP research could remove a major limitation of laser plasma accelerators, a technology that may improve many walks of life.
Episode notes: Flash-X, a multiphysics simulation software package, integrates tools that provide a performance portability solution for exascale computing.
Episode notes: The QMCPACK project intends to provide the capability to find, predict, and control materials and properties at the quantum level.
Episode notes: Machine learning technologies are creating inspiring new opportunity vistas for scientific discovery and research at the exascale.
Episode notes: Project teams can improve the capabilities of math libraries, the foundation of scientific simulations, via cross-project research.
Episode notes: The Extreme-scale Scientific Software Stack (E4S) continues to evolve as a broad collection of software capabilities for scientific research.
Episode notes: An Exascale Computing Project team is getting the longtime open-source PETSc/TAO software suite ready to support exascale applications.
Episode notes: The latest version of the Extreme-scale Scientific Software Stack (E4S) provides support for three different GPU architectures.
Episode notes: ECP plays key support roles in the CANDLE project, which is addressing three significant science challenge problems in cancer research.
Episode Notes: Effective communications, humble problem-solving, lessons learned, and synergy are success factors in realizing the Frontier supercomputer.
Episode Notes: Trilinos, a federated group of software packages with guiding principles, offers much autonomy in solving engineering and science problems.
Episode Notes: Engineer David Grant must check and recheck many important design details before Frontier, the nation’s first exascale supercomputer, is powered on.
Episode Notes: Justin Whitt describes what Frontier will do, why it's unique, progress with deployment, what’s special about exascale computing, and more.
Episode Notes: Doug Kothe explains the need for ECP, the why and how of its origin, its unique team dynamics, and its enduring legacy.
Episode Notes: As principal investigator of the ECP SOLLVE project, Sunita Chandrasekaran emphasizes the development of sustainable software.
Episode Notes: Flux manages and schedules scientific workflows to optimize computing and other resources and enable applications to run faster and better.
Episode Notes: The ExaSGD project wants to enable the day-to-day operation of the national power grid to hold a large number of renewable power sources.
Episode Notes: The latest in the code-for-Aurora series explores an app aimed at high-fidelity whole device modeling of magnetically confined fusion plasmas.
Episode Notes: ExaGraph aims to leave a lasting legacy of algorithms, implementations, and graph-enabled applications for scientific discovery.
Episode Notes: Exascale Computing Project Director Doug Kothe discusses the state of the project, the recent Annual Meeting, and other topics.
Episode Notes: This time, ECP's special podcast series on preparing code for the Aurora exascale supercomputer takes a look at the cosmological code HACC.
Episode Notes: The Exascale Computing Project Software Deployment at Facilities project tests and verifies software functionality and efficiency.
Episode Notes: ECP’s podcast series on preparing code for the Aurora exascale system begins by focusing on an earthquake risk assessment application.
Episode Notes: Simulating earthquake processes from end to end is key to being able to design bridges and buildings to be more resilient to earthquakes.
Episode Notes: ECP's Data and Visualization portfolio is delivering data management software to store, save state, share, and facilitate the analysis of exascale data.
The EXAALT project could bring atomistic materials predictions to the engineering scale and demystify materials design and synthesis.
Episode Notes: Intel executive Trish Damkroger shares her excitement over exascale computing and the design of the Aurora supercomputer.
Episode Notes: Members of the Exascale Computing Project leadership team summarize the state of the project and delve into the major accomplishments.
Episode Notes: Exascale computing could help researchers prepare promising carbon-capture and storage technologies to go from the lab to commercial settings.
Episode Notes: By exploiting the power and performance advantages of GPUs and exascale computing, researchers aim to better predict changes to Earth’s water cycles.
Episode Notes: The Center for Efficient Exascale Discretizations (CEED) recently released version 4.1 of its MFEM software, with features for exascale.
Episode Notes: John Turner of Oak Ridge National Laboratory provides a look at additive manufacturing and the ExaAM project from different angles.
Episode Notes: Spack (Supercomputer PACKage manager) is an R&D 100–Award winner because of its worldwide impact on high-performance computing.
Episode Notes: The ZFP software development effort is tackling the critical task of overcoming the performance cost of data movement for exascale computing.
Episode Notes: Container technology has provided greater software flexibility, reliability, ease of deployment, and portability—an ECP project aims to deliver it for exascale computing systems.
Episode Notes: The WarpX project is developing an exascale application for plasma accelerator research to pave the way for new virtual experiments.
Episode Notes: Machine learning, artificial intelligence, and data analytics are converging with high-performance computing to advance scientific discovery.
Episode Notes: LLVM within the US Department of Energy’s Exascale Computing Project is an open-source collaboration project that develops a compiler infrastructure.
Episode Notes: The Center for Efficient Exascale Discretizations (CEED) is a hub for high-order mathematical methods to increase application efficiency and performance.
The Extreme-scale Scientific Software Stack (E4S) lowers barriers by accelerating the development, deployment, and use of software.
Episode Notes: CODAR, an Exascale Computing Project co-design center, aims to produce an infrastructure for online data analysis and reduction.
Episode Notes: Disciplined and tailored project management led to very impressive results for ECP in a comprehensive independent review.
Episode Notes: DTK and ArborX technologies enable researchers to focus more on their science rather than on low-level algorithms in simulations.
Episode Notes: ECP Earth and Space Science teams are producing sophisticated tools for large-scale scientific discoveries.
Episode Notes: A collaborative team is working to get NWChem ready to run on exascale machines and to provide a starting point for future code development.
Episode Notes: Kathy Yelick and Lenny Oliker of Lawrence Berkeley National Laboratory explain why understanding microbial communities is so important.
Episode Notes: Gina Tourassi discusses the Oak Ridge National Laboratory effort within the Exascale Deep Learning–Enabled Precision Medicine for Cancer (CANDLE) project.
Episode Notes: The ExaSky project conducts extreme-scale simulations to further our understanding of the makeup and evolution of the universe.
Episode Notes: Productivity Sustainability Improvement Planning enables software developers to identify development bottlenecks and track progress to overcome them.
Episode Notes: Successful installation of ECP software requires ensuring speedup for exascale, managing dependencies and versions, and fostering great team communication.
Episode Notes: This discussion examines the xSDK4ECP and hypre projects, which are integrally connected to Exascale Computing Project applications.
Episode Notes: Christian Trott of Sandia National Laboratories shares insights about Kokkos, a programming model for numerous Exascale Computing Project applications.
Episode Notes: Synchronized efforts within the Exascale Computing Project are exploring the use of multiprecision algorithms to achieve faster computation and communication.
Episode Notes: UnifyFS software can provide I/O performance portability for applications, enabling them to use distributed in-system storage and the parallel file system.
Episode Notes: Researchers involved with an input/output software product describe how it will enable users of high-performance computing systems to roll with the rapid pace of change.
Episode Notes: The EXAALT project has made a big step forward with a five-fold performance advance in addressing its fusion energy materials simulations challenge problem.
Episode Notes: The ExaStar project aims to create simulations for comparison with experiments and observations to help answer a variety of questions.
ECP Director Doug Kothe discusses progress assessment and course correction, the consolidation of projects for synergy, ECP's enduring legacy, and more.
Episode Notes: Tom Evans, technical lead for ECP's Energy Applications projects, shares about the motivations, progress, and aspirations on the path to the exascale.
Episode Notes: The work of a project team in lattice quantum chromodynamics is a critical complement to experimentation in particle and nuclear physics.
Episode Notes: Higher resolution, deeper insight, and faster information delivery via exascale computing could vastly improve imaging at X-ray free-electron laser facilities.
Episode Notes: The leader of a project that plans to simulate all aspects of a fusion energy reactor said exascale computing capability will transform the effort.
Episode Notes: The word “ristra” is commonly associated with a string of chile peppers or used to symbolize decorative design. But ECP’s Ristra project is creating a set of computer codes targeting national-security-relevant multiphysics problems.
Episode Notes: Senior computer scientist Franck Cappello leads an effort to compress and reduce the enormous scientific data sets that some of the ECP applications are producing.
Episode Notes: Dave Montoya of Los Alamos National Laboratory and ECP explains aspects of ensuring software in development for exascale is successfully integrated at facilities.
Episode Notes: Cyrus Harrison, Eric Brugger, and Matt Larsen of Lawrence Livermore National Laboratory discuss the first large-scale demonstration of an important supercomputing technology.
Episode Notes: Sameer Shende, an expert on what’s called the Tuning and Analysis Utilities (TAU) performance evaluation tool for applications, explains its features and benefits.
Episode Notes: At the SC18 supercomputing conference in Dallas recently, Salman Habib of Argonne National Laboratory chatted with ECP Communications about cosmological computer modeling and simulation.
Episode Notes: In November at the SC18 supercomputing conference in Dallas, Michael Sprague of the National Renewable Energy Laboratory discussed a recent success of ECP’s wind energy research project, ExaWind.
Episode Notes: The history of supercomputing’s message passing interface (MPI) standard, enhancing asynchronous communication, creating more concurrency at the node level, and the effectiveness of collaboration are among the topics of this episode.
Episode Notes: Mike Heroux and Lois Curfman McInnes discuss how the IDEAS (Interoperable Design of Extreme-Scale Application Software)-ECP helps ease the challenges of software development.
Episode Notes: The software framework called AMReX supports researchers in the study of combustion, cosmology, astrophysics, and accelerators.
Episode Notes: Fred Streitz of Lawrence Livermore National Laboratory discusses his SC18 US Department of Energy booth talk on “Machine Learning and Predictive Simulation: High-Performance Computing and the US Cancer Moonshot on Sierra.”
Episode Notes: A research effort called Datalib: Data Libraries and Services Enabling Exascale Science helps scientists by enabling them to access and share data across platforms, by providing information for troubleshooting, and by building new specialized data services.
Spack (short for Supercomputer PACKage manager) is one of the most popular pieces of software the Lawrence Livermore National Laboratory has ever released to the GitHub open-source community. In ECP, Spack automates the process of releasing versions of math libraries that work together in a single build, thereby making the development of scientific applications better, faster, and cheaper.
Episode Notes: As the need to support hardware accelerators grows, ECP’s Software for Linear Algebra Targeting Exascale (SLATE) project is developing a replacement for the Scalable Linear Algebra PACKage (ScaLAPACK) library, the industry standard for dense linear algebra operations.
Story on ECP’s website: www.exascaleproject.org/converging-an…bra-software/
Episode Notes: Leadership, collaboration, key exascale challenges, and performance measurement are among the topics discussed with ECP Director Doug Kothe.
More on the discussion: https://www.exascaleproject.org/a-discussion-with-ecps-director-august-2018/
Episode 22: The ExaSMR project is advancing modeling and simulation to ensure small modular nuclear reactors will be safe, reliable, and economical. Guest: Steven Hamilton, Oak Ridge National Laboratory.
ExaSMR Project Highlight: https://www.exascaleproject.org/project-highlight-exasmr/
Relatively new within ECP, the Flexible Computational Science Infrastructure (FleCSI) project provides a framework to support multiphysics application development. Guest: Ben Bergen, Los Alamos National Laboratory.
Edited Transcript: https://www.exascaleproject.org/the-flexible-computational-science-infrastructure-flecsi-project-supporting-multiphysics-application-development/
Episode 20: The Exascale MPI project is directed at affecting the evolution of the de facto programming model for parallel computing, message passing interface, so that optimal exascale scientific applications can be developed. Guests: Pavan Balaji and Ken Raffenetti, Argonne National Laboratory.
Edited Transcript: https://www.exascaleproject.org/influencing-the-evolution-of-the-mpi-standard-for-optimal-exascale-scientific-applications/
Computer simulation has become an essential and core component of earthquake design for major infrastructure, but researchers need to better understand and quantify future earthquakes. Exascale computing could allow for the realization of such advances.
Edited Transcript: https://www.exascaleproject.org/exascale-holds-the-key-to-generating-realistic-and-accurate-scenarios-of-future-earthquakes/
The ADIOS project is optimizing I/O on exascale architectures and making itself easily maintainable, sustainable, and extensible, while ensuring its performance and scalability. Guests: Members of the ADIOS research team.
Edited Transcript: https://www.exascaleproject.org/adios-providing-a-framework-for-scientific-data-on-exascale-systems/
ECP’s Software Development Kit effort is an organizational approach to reduce the complexity of the project management of ECP software technology. Guest: Michael Heroux, director of ECP’s Software Technology research focus area.
Edited Transcript: https://www.exascaleproject.org/helping-to-make-the-development-of-scientific-applications-effective-and-efficient-for-exascale-and-beyond/
The Combustion-Pele project uses predictive simulation of in-cylinder combustion processes to explore the potential for groundbreaking efficiences while limiting the formation of pollutants. Guest: Jackie Chen, Sandia National Laboratories.
Edited Transcript: https://www.exascaleproject.org/transforming-combustion-science-and-technology-with-exascale-simulations/
Software called NWChem, or Northwest Chem, can tell researchers a lot about many molecular systems. The NWChemEx project is preparing the software for exascale. Guest: Thom Dunning, the University of Washington and the Northwest Institute for Advanced Computing.
Edited Transcript: https://www.exascaleproject.org/getting-ready-to-model-molecular-properties-at-exascale-with-nwchemex/
The Multiscale Coupled Urban Systems project is applying computing technology and computer models to long-term design and planning issues that pertain to cities. Guest: Charlie Catlett, Argonne National Laboratory.
Edited Transcript: https://www.exascaleproject.org/harnessing-the-power-of-exascale-for-the-long-term-design-and-planning-of-urban-systems/
The Pagoda project aims to develop communication software libraries with low operating overheads that will enable application and software developers to tap the high performance of the US Department of Energy's exascale computers. Guest: Scott Baden, Lawrence Berkeley National Laboratory.
Edited Transcript: https://www.exascaleproject.org/developing-communication-software-libraries-low-operating-overheads-tap-high-performance-exascale-computers/
The xSDK project provides infrastructure and interoperability among numerical libraries developed by diverse groups within ECP. A project that complements xSDK, called IDEAS, is working with the ECP community to understand productivity bottlenecks and to improve both developer productivity and software sustainability, as key aspects of increasing overall scientific productivity. Guest: Lois Curfman McInnes, Argonne National Laboratory.
Edited Transcript: https://www.exascaleproject.org/developing-a-software-ecosystem-for-high-performance-numerical-libraries-and-helping-teams-improve-software-quality/
The Power Steering project within the ECP provides a job-level power management system that can optimize performance under power and/or energy constraints. Guest: Tapasya Patki, Lawrence Livermore National Laboratory
Edited Transcript: https://www.exascaleproject.org/efficiently-using-power-optimizing-performance-scientific-applications-exascale/
The applications of interest to the ECP's Co-Design Center for Particle-Based Applications (COPA) examine the behavior of proteins, materials, astrophysical objects, and more. Guest: Tim Germann, Los Alamos National Laboratory.
Edited Transcript: https://www.exascaleproject.org/ecp-co-design-center-looks-particle-based-applications-exascale/
The Open MPI for Exascale project is focusing on the communication infrastructure of MPI, or message passing interface, an extremely widely used standard for interprocessor communications for parallel computing. Guest: David Bernholdt, Oak Ridge National Laboratory
Edited Transcript: https://www.exascaleproject.org/pushing-mpi-open-mpi-exascale/
The VeloC project endeavors to provide ECP applications an optimal fault-tolerance environment, while the aim of the EZ project is to provide data reduction. Guest: Franck Cappello, Argonne National Laboratory.
Edited Transcript: https://www.exascaleproject.org/addressing-fault-tolerance-and-providing-data-reduction-at-exascale/
The ExaHDF5 project is making HDF5, one of the most popular parallel I/O libraries in high-end supercomputing, ready for exascale. Guests: Suren Byna, Lawrence Berkeley National Laboratory; Quincey Koziol, National Energy Research Scientific Computing Center.
Edited Transcript: https://www.exascaleproject.org/delivering-efficient-parallel-io-with-hdf5-on-exascale-computing-systems/
The Proxy Apps project is curating a collection of proxy apps that will represent the real applications of importance to the ECP. Guest: David Richards, Lawrence Livermore National Laboratory.
Edited Transcript: https://www.exascaleproject.org/picking-best-set-proxy-apps-represent-ecp-applications/
The Application Assessment project was designed to conduct unbiased annual evaluations of the capability, computational performance and scaling, and performance portability of ECP computer codes. Guest: Kenny Roche, Pacific Northwest National Laboratory.
Edited Transcript: https://www.exascaleproject.org/conducting-unbiased-annual-evaluations-of-ecp-applications/
The Exascale Atomistic Capability for Accuracy, Length, and Time (EXAALT) project specializes in molecular dynamics simulations of materials. Guest: Danny Perez, Los Alamos National Laboratory.
Edited Transcript: https://www.exascaleproject.org/extending-reach-molecular-dynamics-simulations-leveraging-exascale/
A conversation with Ames Laboratory’s Mark Gordon on the GAMESS project
Mark Gordon, Ames Laboratory Associate and Distinguished Professor at Iowa State University, spoke with Exascale Computing Project (ECP) Communications at SC17 in Denver. Gordon is the principal investigator for the ECP project called General Atomic and Molecular Electronic Structure System (GAMESS), which is developing methods for computational chemistry. We talked to him about various details of the project and what it has accomplished so far.
Edited Transcript: https://www.exascaleproject.org/making-components-gamess-software-scalable/
A conversation with Los Alamos National Laboratory’s Jim Ahrens on his ECP Data and Visualization project.
Edited Transcript: https://www.exascaleproject.org/understanding-storing-curating-scientific-data-exascale/
Rick Stevens discusses the CANcer Distributed Learning Environment (CANDLE), an ECP application development project targeting new computational methods for cancer treatment with precision medicine.
Edited Transcript: https://www.exascaleproject.org/lighting-way-exascale-precision-medicine/