LLNL Computing’s ground-breaking research and development activities, innovative technologies, and world-class staff are often featured in various media outlets.
Source: Oak Ridge National Laboratory
UnifyFS team wins IPDPS award for open-source software
A research team from Oak Ridge and Lawrence Livermore national labs won the first IPDPS Best Open-Source Contribution Award for the paper “UnifyFS: A User-level Shared File System for Unified Access to Distributed Local Storage.”
Awards Collaborations HPC Systems and Software Open-Source Software Storage, File Systems, and I/O
Source: LLNL Computing
Improving models of clouds and atmospheric turbulence
With simple mathematical modifications to a common model of clouds and turbulence, LLNL scientists and their collaborators helped minimize nonphysical results.
Climate Collaborations Computational Math Computational Science
Source: LLNL News
HPC, AI, and CogSim helped LLNL conquer fusion ignition
The “crystal ball” that provided increased pre-shot confidence in LLNL's fusion ignition breakthrough involved a combination of detailed HPC design and a suite of methods combining physics-based simulation with machine learning—called cognitive simulation, or CogSim.
Source: LLNL News
Visionary report unveils ambitious roadmap to harness the power of AI in scientific discovery
The report lays out a comprehensive vision for the DOE Office of Science and NNSA to expand their work in scientific use of AI by building on existing strengths in world-leading high performance computing systems and data infrastructure.
AI/ML CogSim/AI/ML Hardware Data Science Emerging Architectures Scientific ML
Source: The Next Platform
Talking novel architectures and El Capitan with Lawrence Livermore
LLNL CTO Bronis de Supinski talks about how the Lab deploys novel architecture AI machines and provides an update on El Capitan.
CogSim/AI/ML Hardware Emerging Architectures Exascale HPC Architectures HPC Systems and Software Hybrid/Heterogeneous
Source: LLNL Computing
Investigation of disaggregated memory systems wins poster award
Splitting memory resources in high performance computing between local nodes and a larger shared remote pool can help better support diverse applications.
Awards Data Movement and Memory HPC Systems and Software Hybrid/Heterogeneous
Source: LLNL News
LLNL’s Diachin takes helm of DOE’s Exascale Computing Project
Lori Diachin will take over as director of the DOE’s Exascale Computing Project on June 1, guiding the successful, multi-institutional high performance computing effort through its final stages.
Source: Exascale Computing Project
Podcast: Siting the El Capitan exascale supercomputer
Livermore CTO Bronis de Supinski joins the Let's Talk Exascale podcast to discuss the details of LLNL's upcoming exascale supercomputer.
Emerging Architectures Exascale HPC Architectures HPC Systems and Software Multimedia
Source: YouTube
Video: Data compression for modern supercomputing applications
Unique among data compressors, zfp is designed to be a compact number format for storing data arrays in-memory in compressed form while still supporting high-speed random access.
Data Compression Data Science Multimedia Open-Source Software
Source: YouTube
Video: Vendor-agnostic power management
Variorum provides robust, portable interfaces that allow us to measure and optimize computation at the physical level: temperature, cycles, energy, and power. With that foundation, we can get the best possible use of our world-class computing resources.
HPC Systems and Software Multimedia Open-Source Software Performance, Portability, and Productivity Power Management
Source: LLNL Computing
Best paper winner finds the sources of calculation inconsistencies
Updating a compiler can affect how code runs, leading to inconsistencies in outputs and creating problems for scientists. A new tool automatically finds the sources of these inconsistencies.
Awards Compiler Technology HPC Systems and Software Open-Source Software
Source: LLNL News
LLNL and SambaNova Systems announce additional AI hardware to support Lab’s cognitive simulation efforts
The addition of the spatial data flow accelerator into LLNL’s Livermore Computing Center is part of an effort to upgrade the Lab’s cognitive simulation (CogSim) program.
Co-Design CogSim/AI/ML Hardware Data Science Emerging Architectures HPC Systems and Software Scientific ML
Source: Elastic
National lab uses Elastic to optimize performance of projected world's fastest supercomputer
The Lab was already using Elastic components to gather data from its HPC clusters, then investigated whether Elasticsearch and Kibana could be applied to all scanning and logging activities across the board.
Emerging Architectures HPC Architectures HPC Systems and Software Hybrid/Heterogeneous Information Technology Software Engineering
Source: Data Science Institute
Celebrating the DSI’s first five years
Since 2018, the Lab has seen tremendous growth in its data science community and has invested heavily in related research. Five years later, the Data Science Institute has found its stride.
Source: LLNL News
Girls Who Code ‘Altamont Connection’ brings students from Livermore and Tracy to the Lab
High school students from Livermore and Tracy visited LLNL on May 5 for a first-of-its kind event bringing the two cities’ Girls Who Code (GWC) programs and others together on-site.
Source: LLNL Computing
ISC23 event calendar
LLNL participates in the ISC High Performance Conference (ISC23) on May 21–25.
Co-Design Compiler Technology Data Movement and Memory Emerging Architectures Events Exascale HPC Systems and Software Hybrid/Heterogeneous Power Management Programming Languages and Models Software Build and Installation
Source: Better Scientific Software
Surveys to understand developer health and happiness
Computer scientist Vanessa Sochat talks to BSSw about a recent effort to survey software developer needs at LLNL.
Computational Science Open-Source Software Software Engineering
Source: LLNL Computing
Patent applies machine learning to industrial control systems
A novel ML method discovers and predicts key data about networked devices.
Critical Infrastructure Cyber Data Analytics Cyber Security Data Science Deep Learning Natural Language Processing Network Engineering
Source: Information Week
How do supercomputers fit with strategies for sustainability?
Supercomputers broke the exascale barrier, marking a new era in processing power, but the energy consumption of such machines cannot run rampant.
Exascale HPC Architectures HPC Systems and Software Hybrid/Heterogeneous Power Management
Source: GitHub
Open source is fueling the future of nuclear physics
Open-source software has played a key role in paving the way for LLNL's ignition breakthrough, and will continue to help push the field forward.
Computational Math Computational Science Data Science Discrete Mathematics Lasers and Optics Open-Source Software Scientific ML Transport
Source: LLNL News
Computing codes, simulations helped make ignition possible
For the physicists, computer scientists, and code developers who have worked on fusion for decades, computer simulations have been inexorably tied to the National Ignition Facility’s quest for ignition.
Source: LLNL Computing
A look back: Women in early computing at LLNL
LLNL’s archives recount the contributions of women who developed code during the Lab's early decades.
Careers Diversity HPC Systems and Software Programming Languages and Models
Source: LLNL News
Chandrika Kamath honored as 2023 SIAM fellow
The prestigious fellow designation is a lifetime honorific title and honors SIAM members who have made outstanding contributions to fields served by the organization.
Awards Careers Computational Math Computational Science Data Management Data Science
Source: LLNL News
Scientists develop model for more efficient simulations of protein interactions linked to cancer
The new model addresses a problem in simulating RAS behavior, where conventional methods come up short of reaching the time- and length-scales needed to observe biological processes of RAS-related cancers.
AI/ML Biology/Biomedicine Computational Science Data Science Multimedia
Source: LLNL Computing
Patent for reduced order modeling method boosts lattice design optimization
A new component-wise reduced order modeling method enables high-fidelity lattice design optimization.
Collaborations Computational Math Computational Science Data Science Materials Science Mathematical Optimization