Berkeley Lab Computing Sciences

A comparative infographic showcasing the results of different generative AI models in creating scientific images. The infographic is divided into three vertical sections, labeled 'Ceramics,' 'Plants,' and 'Rocks.' Each section contains a grid of images comparing a 'Raw' scientific source image to AI-generated versions created by models including DCGAN, StyleGAN, DALL-E2, and DALL-E3. The 'Ceramics' section shows microscopic views of packed particles, the 'Plants' section displays images of seedlings and root systems, and the 'Rocks' section features grayscale micro-scans of rock sediment, illustrating the varying success of different AI models in reproducing scientifically accurate images.

Beyond AI art: creating scientific truth.

New @berkeleylab.lbl.gov research evaluates how well AI models generate scientifically valid images of rocks, plants & ceramics. This work could revolutionize research by filling critical data gaps. Learn more: bit.ly/GenAI_LBNL

#AIforScience #GenerativeAI

A digital graphic with a blue circuit board background. On the left is a large circular badge featuring the text “HPCWIRE 2025 Readers’ & Editors’ Choice Awards.” On the right, two images are stacked: above, a scientific visualization with blue and orange isosurfaces; below, a person in a light blue shirt holding a molecular model, set in front of a DNA helix graphic and the word "DOUDNA" written vertically along the left edge.

🎉 Congrats @berkeleylab.lbl.gov researchers for winning 2025 HPCwire Awards!

🏆 @nersc.bsky.social & @VAST: Editor's Choice for advancing scientific data storage powering Doudna #HPC bit.ly/HPCwireDoudna

🏆 HPSF: Readers’ Choice for open software ecosystems bit.ly/HPCwireHPSF

#SC25 #Data #Science

A person wearing a light-colored button-up shirt and a headset microphone gives a presentation while holding a remote control, with geometric red line designs featured in the background.

Congrats to @berkeleylab.lbl.gov 's John Shalf! He’s been awarded IEEE's 2025 Seymour Cray Award for pioneering work in energy-efficient computing & hardware-software co-design bit.ly/JS_CrayAward

Heading to @Supercomputing ( #SC25 ) next week? Connect with our experts: bit.ly/LBL_SC25

#HPC

In a big step for quantum computing, Berkeley Lab and NVIDIA are integrating quantum processors to supercomputers via NVQLink. Our open-source QubiC controller will bring us closer to more powerful and reliable quantum computers. #NVIDIAGTC @atap.lbl.gov
✨ ⬇️
nvidianews.nvidia.com/news/nvidia-...

GenomeOcean Leverages AI to Decode Nature’s Secret Language | Joint Genome Institute Trained on environmental data, a new JGI tool more accurately facilitates microbial genome discoveries.

Meet GenomeOcean: #AI that “speaks” DNA! 🌊🧬 Developed @jgi.doe.gov, this generative AI model is trained on 220+ terabases of #environmental metagenomic data using @nersc.bsky.social #HPC. With 4 billion parameters, it designs novel microbial DNA sequences ➡️ bit.ly/GenomeOcean

@berkeleylab.lbl.gov

A person wearing a formal black suit and tie stands against a plain light-colored background. The individual has light brown hair styled neatly. The image captures the upper body, showcasing a professional appearance suitable for a business or formal setting.

Meet Alex Morehead, @berkeleylab.lbl.gov's 2025 Hopper Fellow! He will develop a foundational #AI model to predict molecular properties & generate new molecules, leveraging @nersc.bsky.social #HPC & open-science collaborations: bit.ly/AM_Hopper

🔗 Apply to be a 2026 #HopperFellow: bit.ly/HopperLBNL

A graphic representation of a coffee cup with layered components illustrating a software architecture. The layers from top to bottom include: 'Compiled Application,' 'PRIF,' 'Caffeine,' and 'GASNet-EX.' The base of the cup is labeled 'System Runtime & Memory Technologies,' indicating the foundational technologies that support the application.

An ASCII art representation of the word 'FIATS' created using various keyboard characters, with the text 'Fiats: Functional Inference and Training for Surrogates' displayed below. The design emphasizes a playful and creative approach to presenting the name of the software framework focused on functional inference and training.

An illustration featuring two interlocking gears labeled 'Julienne' in teal and 'Assert' in dark blue. Surrounding the gears are various design elements, including binary code (010), lines of code, checkboxes, and abstract shapes, symbolizing software development and code reliability. The overall design emphasizes the complementary relationship between the two frameworks in enhancing Fortran code reliability.

DYK @berkeleylab.lbl.gov builds tools to help researchers write faster, more reliable #Fortran —the language powering #supercomputers for decades?

✅ Caffeine: bit.ly/FiatsLBNL

✅ Julienne & Assert: bit.ly/JulienneAssert

✅ Fiats: bit.ly/FiatsLBNL

#AI #HPC #Science #Research #Engineering #Computing

A person wearing a colorful floral-patterned shirt is standing in front of a light blue wall. The shirt features various shades of red, yellow, and green flowers against a dark background. The individual appears to be in a well-lit indoor environment, with soft natural light illuminating the scene.

Unlock the potential of #Fortran, the pioneering programming language still vital today!

Join @berkeleylab.lbl.gov’s Damian Rouson as he dives into Fortran 2023, discussing its significance in #HPC and sharing practical tips for optimizing your research software: go.lbl.gov/hirse-summer...

A graphic celebrating John Shalf's recognition as one of HPCWire's "35 Legends" for 2025. The image features the text "HPC Wire 35 Legends 2025" prominently at the top, with John Shalf's name displayed below it. The caption reads "Champion of HPC-Hyperscale Convergence." The background includes a stylized digital design, suggesting themes of technology and computing.

🎉 Huge congratulations to @berkeleylab.lbl.gov's John Shalf for being named one of #HPCwire’s 35 Legends Class of 2025 — recognized as a Champion of HPC–Hyperscale Convergence!

John’s pioneering work is helping advance scientific computing in the post-Moore’s Law era.

🔗 bit.ly/HPCWireJS

#HPC

An illustration depicting the structure of a massive star at the end of its life. The top half shows the layers of the star: an iron core at the center, surrounded by a silicon-rich layer, and an outer layer labeled as "expelled silicon-sulfur." The bottom half illustrates the supernova explosion, with bright bursts of light radiating outward, indicating the violent event. At the center of this explosion is a neutron star or black hole, representing the remnant core left after the supernova.

In Nature News & Views, @berkeleylab.lbl.gov's Peter Nugent and Harvard CFA’s Anya Nugent discuss a groundbreaking discovery by S.Schulze et al.: the first observation of a silicon- and sulfur-rich layer in a massive star, expelled before #Supernova 2021yfj.

Read more: rdcu.be/eBzj6
#astrophysics

A person seated at a computer desk is focused on their screen while another person stands beside them, pointing at the monitor. The desk is equipped with two large computer monitors displaying various lines of code and data visualizations. The environment appears to be a collaborative workspace, emphasizing teamwork in a technical setting.

DYK #AI + @nersc.bsky.social #HPC can help save lives?

@berkeleylab.lbl.gov researchers developed #ML models to detect health risks earlier. Collaborating with #VA, they analyzed data to uncover key risk factors for suicide and obstructive sleep apnea among U.S. #veterans: bit.ly/4oMPj3y via HPE

Illustration showing a machine learning framework for battery testing: a Gaussian Process model visualizes battery energy predictions across cycles and parameters, with model entropy highlighting areas of uncertainty for new experiments; additional graphics show cell testing results, predicted energy decline, and a gauge indicating battery C-rate from low to high.

Did you know a new #MachineLearning framework developed @berkeleylab.lbl.gov dramatically accelerates #battery lifespan predictions? By combining expert insights with advanced #AI, it delivers accurate results with far fewer experiments—saving time and resources. Learn more: bit.ly/AI_Battery

#HPC

3D simulation of fluid flow through porous dolomite. The rectangular volume shows color gradients from orange and green (indicating fluid movement) on the left, fading to dark blue on the right. White regions highlight high-flow or empty spaces. Simulation data and time step are labeled in the upper left corner.

Ever wondered how fluids twist and loop deep underground?

With @berkeleylab.lbl.gov’s Chombo-Crunch and Dept of Energy’s exascale #supercomputers, scientists can now simulate subsurface flow in unprecedented detail—advancing #geothermal #energy, #water, and #battery research: bit.ly/ChomboCrunch

Basics2Breakthroughs: Simulating pulsars for insights into fundamental physics Formed from collapsed stars, pulsars are tiny astronomical objects that emit stunning, twin beams of light from each pole. Researcher Revathi Jambunathan is developing…

Computational researcher Revathi Jambunathan knows that better understanding pulsars, the spectacular “cosmic lighthouses” of the universe, will lead to better particle accelerators and help us achieve the dream of fusion energy.

Learn more in this video! 🎥 @cs.lbl.gov 🧪

Aydin Buluç: Then and Now / 2013 Early Career Award Winner To solve challenges with ever-larger data sets, Aydin Buluç develops algorithms to increase computers’ efficiency and reduce run time.

A 2013 DOE Early Career Award let Aydin Buluç develop new graph algorithms and scalable computing, speeding discoveries from protein folding to genome assembly.

Now at @berkeleylab.lbl.gov, he advances sparse computations and cross-disciplinary science: bit.ly/LBL_ab

#machinelearning #ai #science

A group of people are gathered around a table in a modern classroom or collaborative workspace, working together on laptops. Some individuals are seated while others are standing and engaging in discussion. In the background, a large screen displays a digital interface, and the room features bright lighting and patterned ceiling panels. The atmosphere appears focused and collaborative.

Six people stand in a row in front of a large screen displaying a news article titled "DOE Announces New Supercomputer Powered by Dell and NVIDIA to Speed Scientific Discovery." The setting appears to be a modern conference or classroom space with glass windows and bright lighting.

@berkeleylab.lbl.gov had a great time teaming up with experts from Livermore Lab and @uhmanoa.bsky.social, sharing supercomputing know-how and exploring advanced energy research with future computational scientists.

Learn more: bit.ly/UHM_lbl

#HPC #supercomputing #energy @nersc.bsky.social

A person standing outdoors in front of tall green trees, wearing a dark crew-neck shirt in natural sunlight.

Meet Mark Fornace, @berkeleylab.lbl.gov's 2025 Alvarez Fellow in Computing Sciences. He’s using advanced #mathematics to unlock new ways for scientists to design and analyze chemical kinetics and #thermodynamics—driving breakthroughs in #bioengineering and beyond.

Learn more: bit.ly/Alvarez_MF

Composite image showing a nuclear physics detector (left) and a supercomputer (right), overlaid on a background of binary code, representing the connection between experimental data collection and high-performance computational analysis.

The DELERIA project is set to revolutionize #NuclearScience! @berkeleylab.lbl.gov & #OakRidgeNationalLab experts are building a data pipeline to connect detectors and supercomputers for real-time analysis—enabling faster insights and discoveries: bit.ly/DELERIA

@esnet.bsky.social @nersc.bsky.social

Five Ways Berkeley Lab's NERSC is Revolutionizing Scientific Research

Five Ways Berkeley Lab's NERSC is Revolutionizing Scientific Research

newscenter.lbl.gov/2025/05/15/f...

#HPC #AI @nersc.bsky.social
@berkeleylab.lbl.gov @cs.lbl.gov

Computational Chemistry Unlocked: A Record-Breaking Dataset to Train AI Models has Launched - Berkeley Lab Scientists will finally be able to simulate the chemistry that drives our bodies, our environment, and our technologies.

DYK Open Molecules 2025—an unprecedented molecular simulation dataset—was just released? Co-led by @berkeleylab.lbl.gov & #Meta, this resource could transform #MachineLearning for real-world #chemistry, #biology, & #energy technologies: bit.ly/OMol25

#AI #MaterialScience #Science #Research #HPC

Presenter holding a microphone stands next to a projection screen displaying a slide titled "The role of simulations in cosmology," which includes diagrams and text about cosmology models, simulations, and mock universes.

Presenter stands in front of a projection screen displaying a slide titled "Random walkers in a potential field," which includes mathematical equations, a 3D plot, and a list of application areas such as chemistry, physics, biology, and social sciences.

Presenter stands at a podium with a microphone in front of a projection screen displaying a simulation visualization of two wind turbines in turbulent atmospheric flow, with a color bar indicating velocity.

Person wearing a red and blue plaid shirt stands at a podium holding a microphone, with a conference badge and lanyard visible. A laptop covered in stickers sits on the podium.

We had an exciting week at the first #HPSF Conference in Chicago! 🌌

@berkeleylab.lbl.gov’s AMReX framework was spotlighted, with 17 app teams sharing breakthroughs in fluid dynamics, turbulent flow, #microelectronics, #AI-enhanced modeling & more.

About our involvement with HPSF: bit.ly/HPSFlbnl

A robotic arm is positioned in a high-tech laboratory setting, surrounded by transparent glass panels. The robot is engaged in materials testing, with various containers organized on a circular platform in the foreground. On the right, shelves hold additional sample jars, highlighting the automated setup designed for efficient scientific research. Bright overhead lights illuminate the workspace.

At @berkeleylab.lbl.gov, we're advancing scientific discovery by integrating #AI, #automation, and data processing. Our #robotic labs, AI-driven lasers, and #supercomputers enhance research efficiency and support reliable energy solutions.

Learn more:
📽️ bit.ly/LBNL_AIvid
📰 bit.ly/LBNL_AI_list

Science Power-Up: Rewriting the Rules with Quantum At Berkeley Lab, scientists are rewriting the rules of computing, chemistry, materials science, and more through the promise of quantum information science. Bert de Jong, Director of the Quantum Systems Accelerator, a DOE National Quantum Information Science Research Center, shares how quantum computers, sensors, and networks are poised to tackle our most complex energy challenges. Collaborating with hundreds of researchers across institutions pushing the limits of hardware and algorithms, quantum research at Berkeley Lab is forging the future of breakthroughs. This episode is the first in our Science Power-Up series, featuring experts across Berkeley Lab on research topics shaping society today, including artificial intelligence, microelectronics, quantum and more. https://www.lbl.gov/research/quantum/ More Berkeley Lab news: http://newscenter.lbl.gov Subscribe: https://youtube.com/berkeleylab Berkeley Lab Social Media: http://socialmedia.lbl.gov

Exciting news! @berkeleylab.lbl.gov's new “Science Power-Up" series starts with a conversation featuring Bert de Jong, Director of the Department of Energy’s Quantum Systems Accelerator.

Learn how we’re harnessing #quantum tech to tackle next-gen science & energy challenges: bit.ly/QSAspu
#Quantum

Taking Big Steps, Berkeley Lab Scientists Shape Improvements in Topological Optimization - Computing Sciences Berkeley Lab scientists have developed a faster method for topological data analysis, improving efficiency in uncovering hidden structures in complex datasets. This breakthrough could transform data a...

DYK @berkeleylab.lbl.gov researchers have developed a faster method for topological #DataAnalysis, significantly improving efficiency in uncovering hidden structures in complex datasets?

This breakthrough has the potential to transform data analysis in fields like #AI & #bio: bit.ly/TopOpt_LBNL

AMCR’s Stefan Wild Named 2025 SIAM Fellow - Computing Sciences Stefan Wild has been named a 2025 SIAM Fellow. This recognition celebrates his groundbreaking work in derivative-free optimization, which has made significant impacts across multiple scientific and en...

🎉 Congrats to @berkeleylab.lbl.gov's Stefan Wild! 🎉

He has been named a 2025 Society for Industrial and Applied Mathematics ( #SIAM ) Fellow for his significant contributions to derivative-free optimization and its applications: bit.ly/WildSIAM25

#AI #hpc #appliedmath #math #engineering #science

A collage of three postdoctoral researchers presenting their work at the Berkeley Lab Computing Sciences Postdoc Symposium. The first researcher is gesturing with their hands while speaking. The second researcher is standing at a podium, explaining their research with hand movements. The third researcher is wearing a black jacket in front of a screen displaying a presentation slide.

Postdocs are the backbone of @berkeleylab.lbl.gov innovations & scientific discovery, playing a pivotal role in our team-science ecosystem.

Learn more about their cutting-edge research from the 2025 Postdoc Symposium. Videos are now available: bit.ly/LBNLcsPDS25

#CS #Datascience #AI #HPC #Quantum

The Perlmutter supercomputer at NERSC features colorful panels with vibrant cosmic imagery, and the name “Perlmutter” is prominently displayed. The panels include images of scientific research and collaboration set against a backdrop of stars and galaxies.

🌟Exciting News!🌟 @berkeleylab.lbl.gov is now an institutional member of the High Performance Software Foundation! This membership advances high-performance #software stacks across industry, academia, and government, meeting the demand for portable and scalable solutions in #HPC + #AI bit.ly/HPSFlbnl

A man with glasses and a beard, wearing a dark blazer and a button-up shirt, sits on a white armchair while speaking and gesturing with both hands. He appears engaged in a discussion on stage, with a gray curtain backdrop and red accent lighting. A small table with a potted plant and water bottles is visible beside him.

⚡️How will next-gen #microelectronics shape the future of #AI?⚡️

Last week, @berkeleylab.lbl.gov’s John Shalf joined experts at the Special Competitive Studies Project’s AI+ Compute & Connectivity Summit to explore the tech driving large-scale AI systems.

🎥 Watch: bit.ly/SCSPaicompute #HPC

#Confab25, @esnet.bsky.social's largest user meeting, will gather domain scientists, network engineers, DOE & NatLab VIPs, ESnet + user-facility staff to co-design the future of integrated science. Register now to strategize & network w/ this cross-cutting group: confab25.es.net/register

A large group of scientists and researchers gather for a group photo in front of Shyh Wang Hall at Berkeley Lab. The participants, diverse in age and background, stand together on a sunny day, smiling at the camera. The modern glass and metal facade of the building serves as the backdrop, with the building name and number "59" visible above the entrance.

A room full of scientists and researchers participate in a collaborative computing event at Berkeley Lab. Attendees sit at tables working on laptops, engaging in discussions, and taking notes. A speaker at the front of the room presents at a podium while two large screens display virtual participants. Others stand along the walls, observing and interacting. The setting is a modern workspace with glass walls and bright lighting.

A group of five people is gathered around a conference table in a meeting room. One person is writing on a whiteboard, explaining a concept to the others. The whiteboard contains mathematical equations and diagrams. Three participants are working on laptops, while two are engaged in discussion. The table has coffee cups, water bottles, and snacks. The setting appears to be a collaborative work session, possibly related to scientific or technical research.

Two women engaged in conversation at a table during a collaborative computing event at Berkeley Lab.

Last week, @berkeleylab.lbl.gov hosted 200 scientists for the 1,000 Scientist #AI Jam, a pioneering event spanning multiple Department of Energy National Labs. They explored how frontier #OpenAI & @anthropic.com models could accelerate breakthroughs in science & nat'l security bit.ly/DOE_LBNL_AIJam