Multi-Scale Simulations Solve a Plasma Turbulence Mystery
Cutting-edge simulations run at NERSC over a two-year period are helping physicists better understand what influences the behavior of the plasma turbulence that is driven by the intense heating necessary to create fusion energy. This research has yielded exciting answers to long-standing questions about plasma heat loss that have previously stymied efforts to predict the performance of fusion reactors and could help pave the way for this alternative energy source. A series of multi-scale simulations run on NERSC’s Edison system found that interactions between turbulence at the tiniest scale (that of electrons) and turbulence at a scale 60 times larger (that of ions) can account for the mysterious mismatch between theoretical predictions and experimental observations of the heat loss. Their findings, published December 17, 2015 in Nuclear Fusion, could greatly improve our knowledge of what’s really going on inside the current tokamak research experiments that exist around the world and in future experimental reactors under construction or planning, the researchers noted.
Who’s New in Computing Sciences?
Each month, Linda Vu profiles new staff and post-docs. Get to know some new faces. »Read February’s batch of new employee profiles.
BIDS Calls for Spring 2016 Data Science Fellows
The BIDS Data Science Fellow Program invites applications for its next cohort of fellows. Successful applicants will join the current cohort of BIDS Data Science Fellows in helping make data analysis easier in the research sciences.
In this call, BIDS seeks applicants in two categories:
- Joint campus appointment
- Graduate student researcher appointment
Each data science fellow will become part of and contribute to a growing ecosystem that brings together faculty, postdoctoral researchers, students, staff, and alumni to form a strong network that assists researchers in advancing data-analysis methods and inquiry, expanding and building new software and analytics tools, sharing best practices& and more.
Potential fellows need not address “big data” to be eligible for the program, but rather be interested in data-related problems that are unique and challenging. In particular, BIDs welcomes applications from cross-disciplinary groups and from individuals exploring topics that may broaden the research diversity of the BIDS community.
This Week’s CS Seminars
Transparent System-Initiated Checkpoint-Restart for High-End HPC
Thursday, March 10, 1:30 to 2:30pm, Wang Hall – Bldg. 59, Room 410
Gene Cooperman, Northeastern University College of Computer and Information Science
The DMTCP project (Distributed MultiThreaded CheckPointing) is now in its second decade. It has a goal of supporting unprivileged, transparent, system-initiated checkpointing across a wide variety of domains. A realistic HPC environment today often interacts with the MPI library itself (e.g., Intel MPI, MVAPICH, or Open MPI), with a Process-Management Interface (e.g., PMI), with a resource manager (e.g., SLURM), and with an RDMA-based fabric (e.g., InfiniBand). In order to transparently adapt to the many possible configurations, DMTCP has replaced the older monolithic checkpointing approach by customizable software layers. DMTCP adapts to the configuration instead of requiring the configuration to adapt to DMTCP. This has now been successfully demonstrated with a “dmtcp module” at an academic supercomputing facility (CCR: Center for Computational Research; State University of New York at Buffalo). Use cases include: chaining a long-running job as separate 24-hour batch jobs; and rolling checkpoints that allow users to scavenge pre-emptible resources. The DMTCP team is also conducting scalability research on Stampede at TACC (with a goal of checkpointing 50,000 CPU cores); and is researching the use of checkpointing for “batch pools” (pools of checkpoint images) instead of traditional batch queues with backfill. This work is joint with Jiajun Cao, and is supported by the National Science Foundation (ACI-1440788) and by a grant from Intel Corporation.
CITRIS Research Exchange: Measuring Regenerative Medical Technology
Wednesday, March 9, 12 to 1pm, Sutardja Dai Hall, 310, Banatao Auditorium
Jan Nolta, UC Davis Health System
Dr. Jan Nolta is the Director of the Stem Cell Program at UC Davis School of Medicine. Her current research is focused on developing therapies that will use mesenchymal stem cells (MSCs) to deliver factors for treating Huntington’s disease and other disorders and injuries. Her group focuses on “bench to the bedside” research, and she has been involved in numerous clinical trials of gene and cell therapy. This talk will address how the field of regenerative medicine can advance into the future of medicine and healthcare hand in hand with devices and wearable technologies to measure outcomes.
This talk is co-sponsored by FEM Tech Berkeley.
Continuous Data Assimilation for Geophysical Models Employing Coarse Mesh Observables
Wednesday, March 9, 3:30 to 4:30pm, 939 Evans Hall, UC Berkeley
Edriss S. Titi, Texas A&M University and Weizmann Institute of Science
One of the main characteristics of infinite-dimensional dissipative evolution equations, such as the Navier-Stokes equations, is that their long-time dynamics is determined by finitely many parameters – a finite number of determining modes, nodes, volume elements and other determining interpolants. In this talk I will show how to explore this finite-dimensional feature of the long-time behavior of infinite-dimensional dissipative systems to design finite-dimensional feedback control for stabilizing their solutions. Notably, it is observed that this very same approach can be implemented for designing data assimilation algorithms for weather prediction based on discrete measurements. In addition, I will also show that the long-time dynamics of the Navier-Stokes equations can be imbedded in an infinite-dimensional dynamical system that is induced by an ordinary differential equation, named determining form, which is governed by a globally Lipschitz vector field. The Navier-Stokes equations are used as an illustrative example, and all the above mentioned results equally hold for other dissipative evolution PDEs, in particular for various dissipative geophysical models.
