NERSC Helps Blaze Path to Next-Gen Lithium-ion Battery
In the quest for a radically better lithium-ion battery, a promising direction is the so-called “lithium-rich” cathode. This kind of battery promises far higher energy densities, which would allow phones and electric vehicles (among other devices) to run for much longer between charges. However, scientists have lacked a clear picture of the chemical processes involved, especially the role of oxygen. Yesterday, a research team led by Gerbrand Ceder of Berkeley Lab’s Materials Sciences Division reported in Nature Chemistry a major advance in understanding how oxygen oxidation creates extra capacity in such cathodes, opening the door to batteries with far higher energy density.
To reach their findings, the team developed a novel methodology of using quantum mechanical simulations to study electron charge transfer in cathode materials with high accuracy. They used supercomputer facilities at the National Energy Research Scientific Computing Center (NERSC), a DOE Office of Science User Facility hosted at Berkeley Lab, and the Extreme Science and Engineering Discovery Environment (XSEDE), led by the University of Illinois.
CS Summer Students Program Kicks off June 7
Computing Sciences welcomes summer students next Wednesday, June 7 with lunch and a talk by Computational Research Division Director David L. Brown. The 12-week program offers undergraduate and graduate students in science and engineering fields the chance to gain research experience with ESnet, NERSC and the Computational Research Division. In addition to completing a research project, the students are given the opportunity to attend weekly talks, tour Berkeley Lab facilities, including NERSC's machine room and the ALS, and present posters outlining their work.
Osni Marques of the Computational Research Division chairs the Summer Student Program.
ESnet Builds Network Model for Maker Faire Crowd
ESnet Interim Director Inder Monga answered questions and demonstrated technologies to attendees of the Bay Area Maker Faire held May 20-22 at the San Mateo County Event Center. Up to 150,000 people were expected to attend the event where the DOE “Make | ENERGY Pavilion” won a “Best in Class” award. Monga led a hands-on demonstration of perfSONAR (PERFormance Service Oriented Network monitoring ARchitecture) built from Raspberry Pi computer kits. The model simulated a worldwide network of servers and participants and used the free and open-source perfSONAR software, developed by ESnet, to troubleshoot bottlenecks. The display was created by Sowmya Balasubramanian, a software developer in ESnet’s Advanced Network Technologies Group, and Mary Hester of ESnet’s Science Engagement Team.
Aspiring Computer Scientists Visit NERSC
A group of 50 enthusiastic computer science students from Dougherty Valley High School in San Ramon, Calif. visited NERSC May 26, where they toured the computer room and participated in lively discussions about the facility and how supercomputers work. They asked great questions, such as "In the future, will there be supercomputers the size of a microchip?" and "What would happen to the supercomputers if there was a huge earthquake or tsunami?" All are AP Computer Science students who have "a deep passion for computer science and science," said their teacher, Preet Dalziel. Most will be majoring in computer science or engineering when they head to college, she added.
This Week's CS Seminars
Friday, June 3
NERSC Data Seminar
Reproducible and Shareable Visualization and Analysis of Mass Spectrometry Imaging Data in OpenMSI with BASTet
12–1 p.m., Bldg 59 Rm 4016
Oliver Rübel, Data Analytics & Visualization Group, CRD, LBNL
Mass spectrometry imaging (MSI) is a transformative imaging method that supports the untargeted, quantitative measurement of the chemical composition and spatial heterogeneity of complex samples with broad applications in life sciences, bioenergy, and health. The development and application of cutting-edge analytical methods is a core driver in MSI research for new scientific discoveries, medical diagnostics, and commercial-innovation. However, the lack of means to share, apply, and reproduce analyses hinders the broad application, validation, and use of novel MSI analysis methods. In this talk we introduce the Berkeley Analysis and Storage Toolkit (BASTet), a novel framework for shareable and reproducible data analysis that supports standardized data and analysis interfaces, integrated data storage, data provenance, workflow management, and a broad set of integrated tools. BASTet serves as the analysis backend for the OpenMSI mass spectrometry imaging science gateway, enabling web-based sharing, reuse, analysis, and visualization of MSI data analyses and derived data products. While this work is motivated by MSI, the challenge to share, reuse, reproduce, and apply novel analyses is common across many application sciences and the methods we describe here are relevant more broadly to this central challenge of modern data science.