Featured Speakers: Anastasiia Butko, Eli Dart, and Zarija Lukić
Date: June 17, 2025
Time: 11 a.m. – 12 p.m.
Where: 59-3101. Virtual: Zoom Available
Three Short Talks:
QC and HPC: the Path from the Experimental to Routine Computing
Presented by Anastasiia Butko (Computer Architecture Group, AMCR)
Quantum computing is a promisingnew technologies that can potentially outperform classical machines for variousscientific applications. Quantum technologies have dramatically evolved overthe past decade, yet it still stays in the area of experimental computing making it challenging for users to adapt for their applications. Moreover, itis unclear how and when quantum accelerators become a part of the largerheterogeneous HPC ecosystem. In this talk, we discuss the path for quantumcomputing the experimental to routine computing.
ESnet: Networking for Science
Presented by Eli Dart (Science Engagement Group, ESnet)
Modern science relies on high performance networks to connect instruments, computing, data, and people together in order to enable new discoveries. This talk will describe science networks as distinct from common ISPs, and provide examples of how large-scale science projects effectively use networking to advance the frontiers of human knowledge.
Simulations and Modeling for Precision Cosmology
Presented by Zarija Lukić (Computational Cosmology Center, SciData)
Ongoing and future cosmology experiments are set to address some of the most profound questions in fundamental science, including the nature of dark energy and dark matter, and testing particle physics and the theory of gravity on the largest observable scales in the universe. However, interpreting the results of these experiments — which span multiple temporal epochs and length scales, presents a formidable challenge. It involves solving an inverse problem: deducing the underlying physics from the acquired observational data. Numerical simulations of the universe provide the most direct approach to tackling this challenge. To meet the challenge, we have developed the Nyx code, which demonstrates remarkable efficiency and scalability. Yet despite these computational successes, creating “virtual universes” remains extremely expensive, especially when high fidelity is required. As a result, our research extends to developing emulators—approximators constructed from a finite set of full-physics simulations, as well as combining physical simulations with generative ML models. These serve as cost-effective surrogates for expensive simulations, allowing us to model scales which would otherwise be out of reach. In this talk I will review the ongoing simulations and modeling effort in the Computational Cosmology Center
Read more about our presenters below.
Featured Speakers: Michael Johnson
Date: June 17, 2025
Time: 1 p.m. – 2 p.m.
Where: Zoom Only
Part of the Nuclear Science Division, the 88-Inch Cyclotron supports ongoing research programs in nuclear structure, astrophysics, heavy element studies, and technology R&D by Lawrence Berkeley National Laboratory (Berkeley Lab) and UC Berkeley. Major instrumentation and facilities at the 88-Inch Cyclotron include the Berkeley Gas-filled Separator (BGS), the Berkeley Accelerator Space Effects (BASE) Facility, and the superconducting VENUS ion source, one of the most powerful Electron Cyclotron Resonance (ECR) ion sources in the world. The 88-Inch Cyclotron Program Head is Dr. Larry Phair.
Virtual tour hosted by Michael Johnson (Research Coordinator, 88-inch Cyclotron, Nuclear Sciences Division)
Featured Speakers: Charles Lively, Lipi Gupta
Date: June 18, 2025
Time: 9 a.m. – 12:30 p.m.
Where: 59-3101. Virtual: Zoom Available
The goal of this training is to introduce new users to NERSC and to provide current NERSC users with a review of using Perlmutter with best practices, such as how to build, execute, and submit jobs on Perlmutter and available tools. Topics covered include Computational systems, Accounts and allocations, Programming environment, Running jobs, Tools, and best practices, and the NERSC data ecosystem.
Training provided by Charles Lively (User Engagement Group, NERSC) and Lipi Gupta (User Engagement Group, NERSC).
Featured Speakers: Charles Lively, Lipi Gupta
Date: June 20, 2025
Time: 9 a.m. – 11:30 a.m.
Where: 59-4102. Virtual: Zoom Available
The goal of this training is to introduce new users to NERSC and to provide current NERSC users with a review of using Perlmutter with best practices, such as how to build, execute, and submit jobs on Perlmutter and available tools. Topics covered include Computational systems, Accounts and allocations, Programming environment, Running jobs, Tools, and best practices, and the NERSC data ecosystem.
Training provided by Charles Lively (User Engagement Group, NERSC) and Lipi Gupta (User Engagement Group, NERSC).
Anastasiia Butko, Ph.D., is a Research Scientist in the Applied Math and Computational Research Division at Lawrence Berkeley National Laboratory (LBNL), California. Her expertise lies in computer architecture, with a focus on high-performance computing (HPC), emerging and heterogeneous technologies, quantum, parallel programming, and architectural simulation techniques. Dr. Butko’s research addresses the challenges of performance scaling in the Post-Moore’s Law era, exploring how alternative technologies can sustain computational advancements.
Eli Dart is a network engineer in the ESnet Science Engagement Group, which seeks to use advanced networking to improve scientific productivity and science outcomes for the DOE science facilities, their users, and their collaborators. Eli is a primary advocate for the Science DMZ design pattern and works with facilities, laboratories, universities, science collaborations, and science programs to deploy data-intensive science infrastructure based on the Science DMZ model.
Zarija Lukić is a staff scientist and a group lead of the Computational Cosmology Center. He worked in the Theoretical Division of Los Alamos National Laboratory before joining Berkeley Lab in 2011. Zarija's research focus is the large-scale structure of the Universe, with the focus on numerical simulations and ML/AI generative methods. An essential component is building simulation codes that can efficiently run on the largest supercomputers, as well as methods for extracting scientific insights using simulation models and observational data from sky surveys.
Charles Lively III, PhD is a Science engagement engineer and HPC Consultant in the User Engagement Group (UEG) under Dr. Rebecca Hartman-Baker at the National Energy Research Scientific Computing (NERSC) Center at Lawrence Berkeley National Laboratory. Charles previously worked at other Department of Energy National Laboratories, including Oak Ridge National Laboratory and Lawrence Livermore National Laboratory.
A Science Engagement Engineer at NERSC, Lipi works to help users engage with NERSC staff and with each other to bridge gaps in scientific computing. She is interested in improving the new user experience, as well as helping graduate students make better use of NERSC resources in preparation for their careers.
