InTheLoop | 03.31.2014
Steve Chu Talk Today
Steve Chu, the former director of Berkeley Lab and former Secretary of Energy, will discuss “Optical Microscopy 2.0” and the energy/climate challenge at 3pm today in the Building 50 Auditorium. A reception will follow.
Human-induced Climate Change Reduces Chance of Flooding in Okavango Delta
Human-induced climate change actually decreased the likelihood of one extreme weather phenomenon seen in recent years: flooding in the Okavango River basin, found a research team that included CRD's Michael Wehner and Daithi Stone computing at NERSC. While the findings seem counter intuitive given the increase in droughts, floods and storms linked to anthropogenic climate change, the researchers say their findings demonstrate that their models are sensitive to natural variations in Earth's climate. This variability manifests itself as 20-30 year periods of above-average or below-average flooding, and the 2009-11 years fall within the former. "If not for climate change, the Okavango system would have experienced even larger flooding in 2009-11 than it actually did,” said University of Cape Town lead author Piotr Wolski. »Read more
CRD’s John Shalf Discusses Big Data with White House Staff
On March 25, John Shalf, head of CRD’s Computer and Data Sciences Department, and three other national laboratory representatives met with President Obama’s chief of staff, members of the Office of Science and Technology Policy (OSTP), Office of Management and Budget (OMB) and other federal agencies to discuss the intersection of big data architecture requirements and exascale challenges. Shalf, who also serves as NERSC's Chief Technologist, was invited to the meeting due to his role as co-leader of the Computer Architecture Lab (CAL), a project to develop energy efficient and effective processor and memory architecture R&D for DOE’s Exascale program. »Read more
ESnet Staff Lead Research Networking Task Forces
Inder Monga and John MacAuley of ESnet helped shape the discussion at the Global Lambda Integrated Facility (GLIF) Technical Working Group meeting held March 19 -20 concurrent in Atlanta. GLIF is an international virtual organization of national research and education networks, consortia and institutions that brings together some of the world's premier networking engineers to develop an international infrastructure. GLIF identifies equipment, connection requirements, and necessary engineering functions and services. Monga lead the GLIF Architecture Task Force which discussed the increasing interest in multilayer and SDN-enabled exchanges. Monga and MacAuley lead the Network Services Interface Task Force.MacAuley lead the AutoGOLE Task Force. »Read more
Reminder: SC14 Technical Paper Abstracts due Friday, April 4
SC14, the premier annual international conference on high-performance computing, networking, storage and analysis, will be held in New Orleans, Louisiana, USA, November 16-21, 2014. The Technical Papers Program at SC is the leading venue for presenting the highest-quality original research, from the foundations of HPC to its emerging frontiers. The Conference Committee solicits submissions of excellent scientific merit that introduce new ideas to the field and stimulate future trends on topics such as applications, systems, parallel algorithms, data analytics and performance modeling. SC also welcomes submissions that make significant contributions to the "state-of-the-practice" by providing compelling insights on best-practices for provisioning, using and enhancing high-performance computing systems, services, and facilities.
Focus areas this year include algorithms, applications, architectures and networks, clouds and distributed computing, data analytics, visualization and storage, performance, programming systems, system software, and state-of-the-practice.
Abstracts due: April 4, 2014 (required)
Full Papers due on April 11, 2014 (up to a one week extension from this date is automatic)
Review Rebuttal Period: June 5-9, 2014
Notification Sent: June 30, 2014
This Week's CS Seminars
DGLM: A high-order discontinuous Galerkin method with Lagrange multipliers for advection diffusion problems
Wednesday, April 2, 3:30pm-4:30pm, 939 Evans Hall, UC Berkeley Campus
Charbel Farhat, Stanford University
A high-order Discontinuous Galerkin method with Lagrange Multipliers (DGLM) is presented for the solution of steady and unsteady advection-diffusion problems in the high Peclet number regime. Unlike HDG methods, it operates directly on the second-order form the advection-diffusion equation and does not require stabilization. Like the Discontinuous Enrichment Method (DEM), it chooses the basis functions among the free-space solutions of the homogeneous form of the governing differential equation, and relies on Lagrange multipliers for enforcing a weak continuity of the approximated solution across the element interface boundaries. For a homogeneous problem, the design of arbitrarily high-order elements based on the proposed method is supported by a detailed mathematical analysis. For a non-homogeneous one, the approximated solution is locally decomposed into its homogeneous and particular parts. The homogeneous part is captured by the DGLM elements designed for a homogenous problem. The particular part is obtained analytically after the source term is projected onto an appropriate polynomial space. An a posteriori error estimator for the proposed method is also derived to enable adaptive mesh refinement. All theoretical results are illustrated by high-order numerical simulations of steady and unsteady problems with steep gradients.
Babbage: the User Friendly Intel Xeon Phi (MIC) Test System at NERSC
Thursday, April 3, 12pm - 1pm, NERSC-OSF room 238
Helen He and Nick Cardo, NERSC
In this talk, we will give an overview of our internal Intel Xeon Phi (MIC) test system, named "Babbage", which is used for application readiness efforts to evaluate new processor architectures. More details at: https://www.nersc.gov/users/computational-systems/testbeds/babbage. Topics to be covered include: the MIC architecture and programming considerations; Babbage system configuration; various tips and efforts making this system very user friendly; how to compile and run; available software and tools; tips and examples of tuning kernels and application performance; users and projects enabled on the system and future plans for Babbage.