InTheLoop | 01.27.2014
Nomination Deadline for Lab Awards Extended to Jan. 31
Nominations are being accepted for the Director's Awards for Exceptional Achievement and The Berkeley Lab Prize-Lifetime Achievement Awards. These awards aim to recognize teams and individuals for significant accomplishments advancing the Lab's mission and strategic goals. Any Lab employee may submit a nomination and honorees will be acknowledged at a Lab awards ceremony and reception. The nomination deadline has been extended to January 31, 2014. The awards are given for science impact, safety, operations, outreach, diversity and several other categories. »Forms and eligibility criteria.
Register Now for NERSC@40 Celebration, NUG Meeting, Feb. 3-6
This year's NERSC Users Group meeting held Feb. 3-6, will mark NERSC's 40th anniversary with a "Celebration of Science and Technology." Two days of talks, reminiscences and the dedication of the Edison system, NERSC's new Cray XC30, will be held Feb. 2-5. Whether you will attend the entire NUG meeting, or simply wish to join the celebration, please register. It's free; it's painless; and it helps us order enough cookies and coffee for the breaks. »See the full agenda and register.
David Skinner Named NERSC Strategic Partnerships Lead
David Skinner has been named to the newly created post, strategic partnerships lead, where he will identify new science communities that can benefit from NERSC resources. In his new role, Skinner, formerly the lead for NERSC's Outreach Software and Programming Group (OSP), will also act as a liaison between NERSC and projects that span many Lab divisions. He will also head a new industrial partnerships initiative, and look for opportunities to collaborate with the Department of Energy's Energy Sciences Network (ESnet) and Berkeley Lab's Computational Research Division. »Read more.
Scientific Computing Profiles CRD's Taghrid Samak
Taghrid Samak of Berkeley Lab's Computational Research Division admits with a laugh that she wasn't one of those kids who started programming on the home computer at age 10. And if she hadn't followed her father's advice, she might have ended up looking for political solutions to pressing problems, rather than working on computational approaches to scientific challenges. »Read more.
A Year of Networking for International Climate Research
This year, ESnet's Science Engagement team will work with the International Climate Network Working Group to setup and optimize network infrastructures for climate data transfers. By the end of 2014, the collaboration hopes to achieve a minimum of 4Gbps data transfer throughput around the globe. »Read more.
Milestone: perfSONAR Deployments Hit 1,000
Pinpointing the problem on large-scale networks can be difficult, so a collaboration of research and academic networking organizations has developed perfSONAR, a publicly available, easy-to-install software suite that takes the guesswork out of network diagnostics. In January 2014, perfSONAR reached a milestone with 1,000 instances of the diagnostic software installed on networking hosts around the U.S. and in 25 other countries. »Read more.
This Week's Computing Sciences Seminars
Low-lying Eigenmodes of the Quark Matrix in Lattice QCD with Exact Chiral Symmetry
Monday, January 27, 2014, 3:00pm - 4:00pm, Bldg. 50F, Room 1647
Ting-Wai Chiu, Professor of Physics, and Associate Director, Center for Quantum Science and Engineering, National Taiwan University
Quantum chromodynamics (QCD) is the fundamental theory for the interaction between quarks and gluons. It manifests as the short-range strong interaction in the nucleus, and plays an important role in the evolution of the early universe, from the quark-gluon plasma phase to the hadron phase. The low-lying eigenmodes of the quark matrix are vital for tackling these nonperturbative physics. In this talk, I review the salient features of the quark matrix in lattice QCD with exact chiral symmetry, and formulate the eigenproblem of the quark matrix, and discuss the strategy to project its low-lying eigenmodes.
SampleClean: Bounding Query Errors in Large and Dirty Datasets
Thursday, January 30, 2014, 2:00pm - 3:00pm, Bldg. 50B, Room 4205
Sanjay Krishnan, University of California, Berkeley
Aggregate query processing over large datasets can be slow and prone to error due to dirty (missing, erroneous, duplicated, or corrupted) values. In this talk, we explore a coupling of sampling and data cleaning. Data cleaning requires either domain-specific software (which can be costly and time-consuming to develop) or human inspection. The latter is increasingly feasible with crowdsourcing but can be highly inefficient for large datasets. Our approach requires cleaning only samples from the dataset to process aggregate numerical queries (e.g., average, sum, count etc). We derive confidence intervals as a function of sample size and show how our approach reduces error bias. We also formulate optimal solutions for cleaning-cost constraints and quality constraints on the queries. We demonstrate our results on benchmarks and real datasets; each of which illustrates different error regimes and their effect on result estimation.
Link of the Week: DOE's 10 Year Exascale Roadmap Crystalizes
At the tail end of 2013, Congress passed a law directing the Department of Energy to develop exascale computing capability within the next decade in order to meet the objectives of the nuclear stockpile stewardship program. In a recent article, HPCWire summarizes progress in that direction. »Read more.