Lawrence Berkeley National Lab

News Archive

2012

Computational Researchers Help Develop Next-Gen Batteries

As part of DOE's new Batteries and Energy Storage Hub, NERSC and CRD resources and expertise will be leveraged to predict the properties of electrolytes. When JCESR is up and running, collaborators will be able to combine these results with the existing Materials Project database to get a complete scope of battery components. Daniel Gunter from ACS is a part of the Materials Project team. Read more ...

ACS Researchers Win Best Paper at Data Cloud Workshop at SC|12

Scientific applications are increasingly using cloud resources for their data analysis workflows. However, managing data effectively and efficiently over these cloud resources is challenging due to the myriad storage choices with different performance, cost trade-offs, complex application choices and complexity associated with elasticity, failure rates in these environments. Devarshi Ghoshal (a summer student) worked with Lavanya Ramakrishnan to explore a framework that uses different data partitioning and distribution strategies to balance application and resource characteristics in cloud environments. The paper describing the initial design and implementation of FRIEDA - a Flexible Robust Intelligent Elastic Data Management framework was awarded Best Paper at the Third International Workshop on Data Intensive Computing in the Clouds (DataCloud 2012). Ghoshal and Ramakrishnan also won the best paper at the workshop last year for their joint work with Shane Canon at NERSC on "I/O Performance of Virtualized Cloud Environments." More details on FRIEDA can be found on the website

Curation of the FLUXNET Global Dataset Leads to AmeriFlux Network Project

Twenty years ago, researchers began installing sensors in a variety of ecosystems to study how carbon dioxide, energy and water vapor cycles through the environment. Today, these sensors have been deployed at 120 locations across the Americas. Because the Department of Energy recognizes that these datasets could benefit a variety of scientific communities, it is funding an effort to make this data accessible to a wide-range of researchers. Margaret Torn of Earth Sciences Division will be leading this effort and will be working closely with ACS member Deb Agarwal. See the full article here.

SC12 Sessions with ACS Contributors

Once again, ACS staff contribute their expertise and experience at the SC12 conference. SC12 will be held Nov 10-16 in Salt Lake City, Utah.

  • Devarshi Ghoshal and Lavanya Ramakrishnan. FRIEDA: Flexible Robust Intelligent Elastic Data Management in Cloud Environments,The Third International Workshop on Data Intensive Computing in the Clouds (DataCloud 2012). Workshop Website
  • Karan Vahi, Ian Harvey, Taghrid Samak, Dan Gunter, Kieran Evans, David Rogers, Ian Taylor, Monte Goode, Fabio Silva, Eddie Al-Shakarchi, Gaurang Mehta, Andrew Jones and Ewa Deelman. A General Approach to Real-time Workflow Monitoring, The 7th Workshop on Workflows in Support of Large-Scale Science (WORKS) 2012. Workshop website
  • Dan Gunter, Shreyas Cholia, Anubhav Jain, Michael Kocher, Kristin Persson, Lavanya Ramakrishnan, Shyue Ping Ong, Gerbrand Ceder, "Accessible Datastore of High-Throughput Calculations: Experiences from the Materials Project", 5th IEEE Workshop on Many-Task Computing on Grids and Supercomputers (MTAGS) 2012 Workshop Website
  • Birds of Feather: HPC Cloud: Can Infrastructure Clouds Provide a Viable Platform for HPC? Kate Keahey, Franck Cappello, Peter Dinda, Dhabaleswar Panda, Lavanya Ramakrishnan. Details
  • SCinet Research Sandbox Presentations: Exploiting Network Parallelism for Improving Data Transfer Performance. Dan Gunter, Raj Kettimuthu, Ezra Kissel, Martin Swany, Jun Yi, Jason Zurawski Details

Two ACS papers will be presented at IEEE eScience Conference

The following two papers were accepted to IEEE eSceience Conference and will be presented in Chicago

  • Taghrid Samak, Dan Gunter and Zhong Wang. Prediction of Protein Solubility in E. coli.
  • Elif Dede, Zacharia Fadika, Jessica Hartog, Madhusudhan Govindaraju, Lavanya Ramakrishnan, Dan Gunter and Richard Shane Canon. MARISSA: MApReduce Implementation for Streaming Science Applications.

ACS Researchers Involved in Two DOE ASCR Collaboratories Projects

The ACS Department recently received funding for work on two collaboratories research projects.

  • Tigres: Template Interfaces for Agile Parallel Data-Intensive Science. This project is addressing the challenge of enabling collaborative analysis of DOE Science data through a new concept of reusable "templates" that enable scientists to easily compose, run, and manage collaborative computational tasks. These templates define common computation patterns used in analyzing a dataset. more. . .
  • FRIEDA: Flexible Robust Intelligent Elastic Data Management. This project investigates an infrastructure strategy that seeks to combine the different usage modalities present in DOE communities under one model that encourages collaborative sharing, supports community growth, accommodates emergent usage patterns such as on-demand computing, and lowers the entry barrier to the use of DOE facilities from desktop to exascale. This project is joint with Argonne National Laboratory. more. . .

Berkeley Lab Hosts Three TechWomen and NERSC Visit By Entire Group

Berkeley lab is once again participating in the TechWomen program. Three emerging leaders from the Middle East and North Africa are visiting the lab for three weeks to work with science mentors at the Advanced Light Source, in Physics, and in Computer Science. TechWomen is an initiative of the U.S. Department of State and matches professionals in the U.S. with counterparts in the program for mentorship and exchange. All forty emerging leaders participating in the program will visit NERSC during the program to learn about high-performance computing and get a tour of the facility. They will also hear talks and interact with a panel of women computer scientists at Berkeley Laboratory.

Berkeley Lab Hosts Week and a half long Discovery Workshops covering High Performance Computing, Cloud Computing and Big Data topics

Co-design, exascale, cloud, big data technologies are evolving and promise to address a number of challenges related to scientific computation and data needs. But cutting through the hype and getting a clear picture on the possibilities and limitations or gaps requires a deep look into the technologies and intensive hands-on sessions. To this end, Berkeley Lab's Computational Research Division (CRD) hosted a weeklong workshop on cloud computing from July 16-20 and a half-week

This year, a number of topics including DOE's Co-Design centers, GPU programming, compiler frameworks, energy efficiency, NoSQL databases and Hadoop were covered. "The workshop was very successful and helped the attendees understand the technologies in depth and separate hype from reality. The expertise at Berkeley Lab in these areas was critical in putting together the content for this workshop," said organizer Lavanya Ramakrishnan of CRD's Advanced Computing for Science Department.

In addition to Berkeley Lab, speakers were from Lawrence Livemore National Lab, UC Berkeley, Intel, Sandia National Lab, Cloudera, Hypertable, Mellanox, IBM Research, Instagram, 10gen, Paradigm4, Yahoo!, Netflix, VMWare, Hewlett Packard. Participation was restricted and the workshop filled up quickly.

LBNL Software Developer Support Brainstorming Sessions

Software has become an essential component of science research. Scientists and engineers throughout LBNL now develop software. Let's brainstorm ways we can build community and support mechanisms for LBNL software developers across the lab. If you are interested in helping, join one of the following sessions:

  • April 4 - noon - 1pm in Perseverance Hall
  • April 5 - 3pm - 4pm at OSF (238)
  • April 9 - 10am - 11am in Perseverance Hall
Organizers: Deb Agarwal (daagarwal@lbl.gov), David Skinner (deskinner@lbl.gov), and Eric Pouyoul (epouyoul@lbl.gov).

Conference on Data Analysis Keynote and Posters

The recent "Conference on Data Analysis: Exploring Research across the Department of Energy", featured Deb Agarwal as a keynote speaker on the topic "Changing Your Perspective from Serving the Data to Enabling Data Users." Taghrid Samak and Arthur Weidmer were also featured at the conference as invited poster presenters.

2011

Microsoft Research Website Highlights ACS Research

Microsoft Research recently posted three science stories involving collaborations with the Computational Research Division's (CRD's) Advanced Computing for Science Department (ACS) and the Berkeley Water Center (BWC):

The ACS team involved in the work includes Deb Agarwal, Monte Goode, Keith Jackson, and Gary Kushner. They collaborated closely with BWC personnel, Marty Humphrey and Norm Beekwilder of University of Virginia, and Catharine van Ingen of Microsoft Research on these projects.

ACS/NERSC Summer Students Score a Hat-Trick

It is rare to submit three papers to the same conference, but even rarer to have all of them accepted! But this is exactly what happened in the summer of 2011 when two summer students, Zach Fadika and Elif Dede, had all three of their submitted papers accepted at the Grid2011 conference held in Lyon, France. Elif Dede was advised by Dan Gunter and Lavanya Ramakrishnan from ACS, and Zach Fadika was also advised by Lavanya Ramakrishnan with Shane Canon from NERSC.

The three "hat-trick" papers were:
  1. Benchmarking MapReduce Implementations for Application Usage Scenarios. Zacharia Fadika, Elif Dede, Madhusudhan Govindaraju and Lavanya Ramakrishnan
  2. MARIANE: MApReduce Implementation Adapted for HPC Environments. Zacharia Fadika, Elif Dede, Madhusudhan Govindaraju and Lavanya Ramakrishnan
  3. Scalable and Distributed Processing of Scientific XML Data. Elif Dede, Zacharia Fadika, Chaitali Gupta and Madhusudhan Govindaraju

We look forward to hearing many more good things from Zach and Elif (and, indeed, all our summer students) in the future.

ACS Researchers Lead UQ Study Group

Mathematical models intended for computational simulation of complex real-world processes are a crucial ingredient in virtually every aspect of DOE science. Utilization of such computer models requires addressing Uncertainty Quantification (UQ), a greatly expanding field focusing on systematic uncertainties in computer models such as model limitations, sparse inputs, and initial conditions; and statistical uncertainties such as noisy or incomplete data. Within LBNL and CRD in particular, there is an excellent mix of skills in mathematics, modeling, parallel programming, and systems to address these challenges head-on. But current CRD approaches to understanding and using UQ are scattered across researchers and departments.

To help better focus CRD efforts, Dan Gunter and Lavanya Ramakrishnan of the Advanced Computing for Science (ACS) Department formed an Uncertainty Quantification Study Group that will help familiarize participants with the varied aspects of the UQ challenge, and spur discussions of how CRD research can be applied to their solution. The study group is an opportunity to cross-germinate ideas and explore collaborations across groups. Read more on the UQ group wiki..

Berkeley Lab Hosts Week and a half long Discovery Workshops covering HPC and Cloud Computing topics

Exascale computing and cloud computing are both rapidly emerging as key capabilities available to support scientific computation. The forecast for many aspects of computing definitely calls for exascale computing or clouds, but cutting through the hype and getting a clear picture on the possibilities and limitations of each requires a deep look into the technologies and intensive hands-on sessions. To this end, Berkeley Lab's Computational Research Division (CRD) and NERSC hosted a weeklong workshop on cloud computing from June 20-24 and a half-week workshop on high-performance computing advances June 15-17.

The Center for Information Technology Research in the Interest of Society (CITRIS)at UC Berkeley helped host the workshop. Participating organizations helping Berkeley Lab provide content include UC Berkeley, Google, Yahoo!, Amazon, Cloudera, the and Microsoft. The Department of Energy's National Energy Research Scientific Computing Center (NERSC) contributed workshop content. Participation was limited and the limit was quickly met.

"There is a lot of information floating around on exa-scale and cloud computing, but it's not easy to separate fact from fiction. Fortunately, we have a number of research projects, testbeds and collaborations working in these areas, so we were in a good position to marshal the necessary resources for this workshop," said organizer Keith Jackson, of CRD's Advanced Computing for Science Department.

2010

Berkeley Lab Contributes Expertise to New Amazon Web Services Offering

This month Amazon Web Services (AWS) lauched a new Cluster Compute Instances offering for Amazon EC2, which will make high-bandwidth, low-latency HPC resources available in a cloud-computing environment. To ensure that this new service can handle a gamut of demanding HPC applications, AWS staff worked closely with researchers in the Berkeley Lab's Computational Research Division (CRD), Information Technologies Division, and NERSC.  read more..

- Linda Vu