-->
Skip to Main Content U.S. Department of Energy
Lecture Series

Featured Speakers

2017

Scalable Tensor Decompositions in High Performance Computing Environments

JLi
Dr. Jiajia Li

December 14, 2017
Location: BSF Crick Room (2008)
Time: 10 AM
Presented by: Dr. Jiajia Li, Ph.D. Candidate, Computational Science and Engineering, Georgia Institute of Technology

Dr. Jiajia Li will present scalable tensor decompositions in high-performance computing platforms. Tensor decompositions are noted for their ability to discover multidimensional dependencies and have found numerous uses in healthcare analytics, social networks analytics, machine learning, etc. Dr. Li will propose different techniques to overcome challenges in using tensor decompositions, including the memorization method. This Computing@PNNL Seminar is being hosted by Shuaiwen Leon Song, Physical and Computational Sciences Directorate.
View Seminar Series Flier


The DOE Exascale Computing Initiative: ECP and Holistic Co-design

JAA
Dr. James A. Ang

December 12, 2017
Location: BSF Crick Room (2008)
Time: 10 AM
Presented by: Dr. James A. Ang, Technical Manager, Exascale Computing Program, Sandia National Laboratories

Dr. James Ang will describe the Department of Energy’s role under the structure of the National Strategic Computing Initiative (NSCI) and the differences between the Exascale Computing Initiative (ECI) and the Exascale Computing Project (ECP). He will include an overview of ECP and holistic co-design, as well as touch on other NSCI/ECI activities relevant to potential interagency collaborations in advanced architectures. This Computing@PNNL Seminar is being hosted by Nathan Baker, Physical and Computational Sciences Directorate.
View Seminar Series Flier


A Class of HPCC Framework for Uncertainty and Sensitivity Analysis

DG
Dr. Devan Ganesh

November 21, 2017
Location: ISB2 Wanapum Room (155)
Time: 10 AM
Presented by: Dr. Devan Ganesh, Professor, College of Engineering & Computational Sciences, Department of Applied Mathematics & Statistics/Computer Science Department, Colorado School of Mines

This talk will consider a class of wave propagation models with aleatoric and epistemic uncertainties. Using mathematically robust parameter estimates, Professor Ganesh will describe the development of high performance and cloud computing (HPCC) offline/online strategies to compute statistical moments of a key quantity of interest in such models. He also will focus on an HPCC framework, a hybrid of reduced order models (ROM) and HPC, that can be used in conjunction with any computational method to simulate a single particle deterministic wave propagation model. This Computing@PNNL Seminar is being hosted by Nathan Baker, Physical and Computational Sciences Directorate.
View Seminar Series Flier


Image Deblurring and Video Stabilization of Turbulence Distortion

YLou
Dr. Yifei Lou

November 14, 2017
Location: CSF Darwin Room (1007)
Time: 10 AM
Presented by: Dr. Yifei Lou, Assistant Professor of Mathematics, The University of Texas at Dallas

This talk will focus on “image deblurring,” which often involves solving a numerically ill-conditioned backward-diffusion problem. Professor Lou will detail a direct deblurring method using an explicitly generative model together with sparse characteristics of natural images. With it, a sparse representation of the blurred image can be computed and its coefficients used to combine elements of the original basis, leading to a restored image. She also will examine “imaging through turbulence” and a method that combines Sobolev gradient and image diffusion to reduce the artifacts of blurring and oscillation simultaneously. This Computing@PNNL Seminar is being hosted by Xiu Yang, Physical and Computational Sciences Directorate.
View Seminar Series Flier


Recent Advances in Gradient-Based Damage Methods for Fracture and Fragmentation

JED
Dr. John E. Dolbow

November 06, 2017
Location: ISB1 White Bluffs Room (105)
Time: 11 AM
Presented by: Dr. John E. Dolbow, Professor, Civil and Environmental Engineering, and Director, Computational Mechanics Laboratory, Duke University

Professor Dolbow will discuss a recent method for fracture and fragmentation that builds on other established links between gradient-based damage models and cohesive models of failure. He also will present results for a series of benchmark problems in large-scale dynamic fracture and fragmentation and highlight progress on extended finite element methods to transition from regularized to discontinuous representations of failure surfaces. This Computing@PNNL Seminar is being hosted by Nathan Baker, Physical and Computational Sciences Directorate.
View Seminar Series Flier


A Game-Theoretic Approach to Numerical Approximation and Algorithm Design

DHO
Dr. Houman Owhadi

October 30, 2017
Location: ISB2 Wanapum Room (155)
Time: 10 AM
Presented by: Dr. Houman Owhadi, Professor of Applied and Computational Mathematics and Control and Dynamical Systems, California Institute of Technology, Engineering and Applied Science

This talk will examine the interplay between game theory, numerical approximation, and Gaussian process regression. Through examples, Professor Owhadi will illustrate the interface between statistical inference and numerical analysis via multiscale analysis problems, fast solvers design, operator adapted wavelet identification, and computations with dense kernel matrices. This Computing@PNNL Seminar is being hosted by Xiu Yang, Physical and Computational Sciences Directorate.
View Seminar Series Flier


Integrated Computational and Experimental Materials Engineering (ICEME) for Vehicle Lightweighting

RKM
Dr. Raja K. Mishra

September 07, 2017
Location: ISB2 Wanapum Room (155)
Time: 9:30 AM
Presented by: Dr. Raja K. Mishra, Technical Fellow, General Motors Research & Development Center

Computational materials engineering tools capable of integrating microstructure-based material and process design with performance-driven structural optimization can significantly enhance manufacturing competitiveness. Thus, the automotive industry is embracing Integrated Computational and Experimental Materials Engineering, or ICEME, tools. Dr. Mishra’s talk will present multiscale computational frameworks involving coupled micro- and macro-scale numerical models for high strength aluminum alloys, advanced high strength steels, and magnesium alloys. This Computing@PNNL Seminar is being hosted by Xiaohua Hu, Physical and Computational Sciences Directorate.
View Seminar Series Flier


Capturing Provenance as a Diagnostic Tool for Workflow Performance Evaluation and Optimization

LP
Dr. Line Pouchard

August 31, 2017
Location: ISB1 White Bluffs Room (105)
Time: 1 PM
Presented by: Dr. Line Pouchard, Senior Researcher, Computational Science Initiative Directorate, Center for Data Driven Discovery (C3D), Brookhaven National Laboratory

In her talk, Dr. Pouchard will demonstrate how provenance is a useful tool for evaluating and optimizing workflow performance in extreme-scale high-performance computing environments. From the Exascale Computing Project CODAR project, she also will present Chimbuko, a framework for analysis and visualization of the provenance of workflow performance that enables exploration of performance metrics data. This Computing@PNNL Seminar is being hosted by Eric Stephan, Physical and Computational Sciences Directorate.
View Seminar Series Flier


Predictive Assessment of the Human and Environmental Toxicity of Chemical Substances

GS
Dr. Gerrit Schüürmann

August 25, 2017
Location: CSF Darwin Room (1007)
Time: Noon
Presented by: Prof. Dr. Gerrit Schüürmann, Head, Department of Chemical Ecotoxicology, UFZ Centre for Environmental Research, and Professor, Technical University of Freiberg

In his talk, Prof. Dr. Gerrit Schüürmann will present the integrated testing strategy, ITS, and respective non-test (in silico) and test (in chemico; in vitro) methods for alternative toxicological assessment of chemical compounds, as well as the roles of chemoavailability and the exposome platform for a molecular-level understanding of chemical toxicity. This Computing@PNNL Seminar is being hosted by Robert Rallo, Physical and Computational Sciences Directorate.
View Seminar Series Flier


Mechanics of Cell Interactions with Low-Dimensional Nanomaterials

HG
Dr. Huajian Gao

July 19, 2017
Location: BSF Crick Room (2008)
Time: 11:30 AM
Presented by: Dr. Huajian Gao, Walter H. Annenberg Professor of Engineering, Brown University

In his talk, Professor Huajian Gao, of Brown University, will examine cell interactions with low-dimensional nanomaterials. Nanomaterials, including various types of nanoparticles, nanowires, nanofibers, nanotubes, and atomically thin plates and sheets, are being produced and released into the environment at a rate of nearly a million tons per year, making it critical to understand the impact of these biological interactions. This Computing@PNNL Seminar is being hosted by Nathan Baker, Physical and Computational Sciences Directorate.
View Seminar Series Flier


Trustworthy Data: Multiparty Data Vaults and Data Collaboratives in Justice, Smart Cities and Critical Infrastructure

EDM
Brant Zwiefel

June 23, 2017
Location: ISB2 Wanapum Room (155)
Time: 10:30 AM
Presented by: Brant Zwiefel, Managing Architect and Worldwide Innovation Lead Public Sector and Industries, Microsoft

In his talk, Microsoft’s Brant Zwiefel will will examine the complexities affecting modern information technology systems, specifically at how organizations are leveraging technology innovations like the mobile, social, cloud, and big data to be more connected through systems of engagement that enable organization to better model their environment. Multiparty Data Vaults and Data Collaboratives provide a platform for secure data sharing that span both functions within an organization, as well as across organization boundaries while maintaining data provenance and chain of custody necessary to create, maintain, and enhance trust. This Computing@PNNL Seminar is being hosted by Robert Rallo, Physical and Computational Sciences Directorate.
View Seminar Series Flier


Application of Computational Methods in Support of Steel Grade Development

EDM
Dr. Emmanuel De Moor

June 22, 2017
Location: ISB2 Wanapum Room (155)
Time: 10 AM
Presented by: Dr. Emmanuel De Moor, Associate Professor, George Ansell Metallurgical and Materials Engineering Department, Colorado School of Mines

Professor Emmanuel De Moor, of Colorado School of Mines, will explore the need to develop higher-strength steel grades with adequate formability, primarily for the automotive industry, which is continually challenged to achieve vehicle lightweighting without sacrificing occupant protection. His talk will include approaches to alloying, processing, and microstructure formation, as well as applications of computational tools and methods working toward accelerated steel development. This Computing@PNNL Seminar is being hosted by Xiaohua Hu, Physical and Computational Sciences Directorate.
View Seminar Series Flier


Optimizing Recursive Task Parallel Programs

VKN
Dr. V.K. Nandivada

June 12, 2017
Location: CSF Darwin Room (1007)
Time: 1 PM
Presented by: Dr. V. Krishna Nandivada, Associate Professor, India Institute of Technology Madras, Department of Computer Science and Engineering

Professor V.K. Nandivada, of IIT Madras, will present a new optimization DECAF that optimizes recursive task parallel programs by reducing the task creation and termination overheads. This DECAF has been implemented in the X10v2.3 compiler and tested over a set of benchmark kernels on two different hardwares, a 16-core Intel and 64-core AMD system, with both achieving considerable speedup. This Computing@PNNL Seminar is being hosted by Sriram Krishnamoorthy, Physical and Computational Sciences Directorate.
View Seminar Series Flier


Multiscale Mechanics and Materials Research for Energy Efficiency and Extreme Environments

HZ
Dr. Hussein M. Zbib

April 20, 2017
Location: CSF Darwin Room (1007)
Time: 9:30 AM
Presented by: Dr. Hussein M. Zbib, Professor, Washington State University School of Mechanical and Materials Engineering

Professor Hussein Zbib, of WSU, will provide an overview of research activities focused on a multiscale experimental and computational predictive capability that enables fundamental insights into the performance of new classes of materials with superior properties. This Computing@PNNL Seminar is being co-hosted by Kyoo Sil Choi and Xiaohua Hu, Physical and Computational Sciences Directorate.
View Seminar Series Flier


Performance Advantages of a Data-Centric Computer Architecture

Ken
Dr. Kenneth P. Jacobsen

March 23, 2017
Location: ISB2 Wanapum Room (155)
Time: 10 AM
Presented by: Dr. Kenneth P. Jacobsen, Chief Executive Officer, EMU Technology

Dr. Ken Jacobsen, CEO of EMU Technology, will focus on data-centric architectures and introduce a new, highly-scalable partitioned global address space (PGAS) architecture. The PGAS architecture can scale to very large sizes using a shared memory programming model that is relatively invisible to the programmer. In this architecture, the program thread migrates to the data, and data never move, which can result in dramatic performance improvements. This seminar is being hosted by Nathan Baker, Physical and Computational Sciences Directorate.
View Seminar Series Flier


Uncertainty Quantification in Computational Models

Najm
Dr. Habib N. Najm

March 16, 2017
Location: CSF Mural Room (1508A)
Time: 10 AM
Presented by: Dr. Habib N. Najm, Distinguished Member, Technical Staff Combustion Research Facility, Reacting Flow Research, Sandia National Laboratories

Dr. Habib Najm, from Sandia, will discuss recent uncertainty quantification, or UQ, developments including the statistical estimation of model error, which is useful for assessment of meaningful predictive uncertainties and for diagnostic purposes in model development. He will examine using multi-level, multi-fidelity methods in the context of both global sensitivity analysis and forward UQ to address the challenge of high dimensionality and computational complexity and will illustrate using these capabilities in the context of targeted model problems. This Frontiers Lecture is being hosted by Alexandre Tartakovsky, Physical and Computational Sciences Directorate.
View Seminar Series Flier


The Center for Expanded Data Annotation and Retrieval: Making Data ‘FAIR’

Musen
Dr. Mark A. Musen

March 13, 2017
Location: BSF Crick Room (2008)
Time: 11 AM
Presented by: Dr. Mark A. Musen, Professor, Biomedical Informatics and Biomedical Data Science, and Director, Stanford Center for Biomedical Informatics Research, Stanford University

Dr. Mark Musen, from Stanford University, will discuss his work as the Principal Investigator of Stanford’s Center for Expanded Data Annotation and Retrieval (CEDAR), a center of excellence in the NIH Big Data to Knowledge Program with the goal of enhancing the authoring of experimental metadata to make online data sets more findable, accessible, interoperable, and reusable—or FAIR. He will explain the FAIR principles and how CEDAR’s work may ease access to and reuse of scientific data sets stored in public repositories. This Frontiers Lecture is being hosted by Nathan Baker, Physical and Computational Sciences Directorate.
View Seminar Series Flier


Uncertainty Quantification in Multiscale Simulations

Zabaras
Dr. Nicholas Zabaras

February 10, 2017
Location: ISB2 Wanapum Room (155)
Time: 10 AM
Presented by: Dr. Nicholas Zabaras, Viola D. Hank Professor of Computational Science and Engineering, University of Notre Dame

Dr. Nicholas Zabaras, from Notre Dame, will address two unifying themes typical of multiscale simulations using materials physics as the application domain: 1) the use of surrogate models and 2) stochastic coarse graining (CG). He will advocate data-driven machine learning approaches for both challenges. His talk will examine a Bayesian framework for alloy modeling using the cluster expansion, where he will show the framework’s performance in predicting various thermodynamic properties and will present an assessment of a data-driven, CG atomistic formulation in the context of equilibrium statistical mechanics. This Frontiers Lecture is being hosted by Nathan Baker, Physical and Computational Sciences Directorate.
View Seminar Series Flier


Towards a Quantum Future of Computation

Troyer
Dr. Matthias Troyer

January 27, 2017
Location: BSF Crick Room (2008)
Time: 11 AM
Presented by: Dr. Matthias Troyer, Principal Researcher, Microsoft Quantum Architectures and Computation Group

Dr. Matthias Troyer, currently with Microsoft (on leave from the Swiss Federal Institute of Technology ETH Zurich), will examine the roadmap to building universal quantum computers with important real-world applications. He will explain why a quantum computer needs to be not only asymptotically superior but able to solve problems within a limited time that no classical supercomputer can solve and will touch on the steps of quantum software engineering needed to turn a quantum algorithm into a “quantum killer app.” He also will review how quantum algorithms improvements have turned problems in materials science and quantum chemistry to realistic applications of quantum computers with varied research applications. This Frontiers Lecture is being hosted by Nathan Baker, Physical and Computational Sciences Directorate.
View Seminar Series Flier


Integrated Computational Materials Engineering

KM
Dr. Karel Matouš

January 10, 2017
Location: CSF Darwin Room (1007)
Time: 11 AM
Presented by: Dr. Karel Matouš, College of Engineering Collegiate Associate Professor of Computational Mechanics and Director, Center for Shock Wave-processing of Advanced Reactive Materials (PSAAP II), University of Notre Dame

Dr. Karel Matouš, from Notre Dame, will present an image-based (data-driven) multiscale framework for modeling the chemo-thermo-mechanical behavior of heterogeneous materials while capturing the large range of spatial and temporal scales. This integrated computational approach for predicting the behavior of complex heterogeneous systems combines macro- and micro-continuum representations with statistical techniques, nonlinear model reduction, and high-performance computing that aims to benefit efforts to develop new multifunctional materials. This Frontiers Lecture is being hosted by Nathan Baker, Physical and Computational Sciences Directorate.
View Seminar Series Flier



2016

Modeling the Effects of Cure-induced Chemo-mechanical Processes on the Strength of Fiber-reinforced Composites

Waas
Dr. Anthony M. Waas

August 26, 2016
Location: CSF Darwin Room (1007)
Time: 10 AM
Presented by: Dr. Anthony M. Waas, Boeing-Egtvedt Chair, Professor of Aerostructures/Chair, Department of Aeronautics and Astronautics, University of Washington

Dr. Anthony M. Waas, from the University of Washington, will introduce a novel computational model to analyze the effect of the curing process on the in-service performance of fiber-reinforced composite structures. He also will present examples from continuous fiber-reinforced polymer matrix and textile composites. This Frontiers Lecture is being hosted by Xin Sun, Physical and Computational Sciences Directorate.
View Lecture Series Flier



2015

Nanoscale Heat Conduction: From Molecular Dynamics to Nonlocal Models and Nonlocal Models to Local Models

Du
Dr. Xiantao Li

June 26, 2015
Location: ISB2/Wanapum Room (155)
Time: 10 AM
Presented by: Dr. Xiantao Li, Associate Professor, Department of Mathematics, The Pennsylvania State University

Dr. Xiantao Li, from Penn State University, will present a mathematical derivation of a heat conduction model. The derivation begins with full classical molecular dynamics, leading to a nonlocal model—both in space and time. He also will discuss how to further simplify the nonlocal model by introducing additional equations for the heat flux, particularly by deriving a hierarchy of heat conduction models. This Frontiers Lecture is being co-hosted by Adolfy Hoisie, Fundamental & Computational Sciences Directorate, and Nathan Baker, National Security Directorate.
View Lecture Series Flier


Nonlocal Calculus of Variations and Asymptotically Compatible Schemes

Du
Dr. Qiang Du

June 22, 2015
Location: BSF Crick Room (2008)
Time: Noon
Presented by: Dr. Qiang Du, Fu Foundation Professor of Applied Mathematics, Columbia University

Dr. Qiang Du, from Columbia University, will present recent developments in nonlocal calculus of variations and asymptotically compatible schemes. These methods have provided a framework for mathematical understanding of nonlocal operators and nonlocal models. Moreover, they may be important for validation and verification of multiscale models and simulations. This Frontiers Lecture is being co-hosted by Adolfy Hoisie, Fundamental & Computational Sciences Directorate, and Nathan Baker, National Security Directorate.
View Lecture Series Flier


The Role of the Cloud and Big Data Analytics in Scientific Research and the Search for Scalable Tools

Gannon
Dr. Dennis Gannon

March 23, 2015
Location: CSF Mural Room (1508A)
Time: 10:00 AM
Presented by: Dr. Dennis Gannon, Director of Cloud Research Strategy, Microsoft Research (retired)

Dr. Dennis Gannon, former Director of Cloud Research Strategy with Microsoft Research, will showcase examples of how science is evolving because of cloud and machine learning advances and will discuss some of the challenges that must be addressed to make tools truly available to the broader scientific research community. This Frontiers Lecture is being co-hosted by Adolfy Hoisie, Fundamental & Computational Sciences Directorate, and Nathan Baker, National Security Directorate.
View Lecture Series Flier


Data Analytics, Exascale Architectures and Computer Science:
The Path to Tomorrow

Hendrickson
Dr. Bruce Hendrickson

February 06, 2015
Location: EMSL Auditorium
Time: 10:00 AM
Presented by: Dr. Bruce Hendrickson, Director of Computing Research at Sandia National Laboratories

In his talk, Dr. Bruce Hendrickson, of Sandia National Laboratories, will explain how fundamental physical and engineering constraints are creating difficult programming challenges and adding to the complexities facing next-generation computing. He will survey these complex trends and attempt to glimpse the road ahead, including sharing insights on how the scientific computing and data science communities have interesting ideas for improving software productivity that may be leveraged to move technology forward. This Frontiers Lecture is being co-hosted by Adolfy Hoisie, Fundamental & Computational Sciences Directorate, and Nathan Baker, National Security Directorate.
View Lecture Series Flier


Advanced Potential Energy Surfaces for Condensed Phase Simulation

Teresa Head-Gordon
Dr. Teresa
Head-Gordon

January 26, 2015 Location: ETB Columbia River Room
Time: 10:00 am
Presented by: Dr. Teresa Head-Gordon Professor, Departments of Chemistry, Bioengineering, and Chemical and Biomolecular Engineering, University of California, Berkeley Scientist, Lawrence Berkeley National Laboratory, Berkeley, California

Dr. Head-Gordon's research program focuses on the interface of theoretical chemistry with biology and physics. In her talk, Dr. Head-Gordon will introduce new theoretical models and methods that include direct and mutual polarization based on the AMOEBA (atomic multipole optimized energetics for biomolecular applications) polarizable force field and a Poisson-Boltzmann semi-analytic method (PB-SAM). These models and their implementations on multicores are opening new abilities to allow larger scales of study for molecular simulation with more complex potential energy surfaces.

A professor at UC Berkeley and scientist at Lawrence Berkeley National Laboratory since 1992, Dr. Head-Gordon also serves as a faculty staff scientist at the California Institute for Quantitative Biosciences and at Clare Hall, Cambridge, United Kingdom. She has received the IBM Shared University Research (SUR) Award in 2001 and served as the Schlumberger Professor, Cambridge University, United Kingdom in 2005-2006. Dr. Head-Gordon was a panel member of the U.S. National Academies Study on potential impact of advances in high-end computing in science and engineering in 2006-2007 and of the National Institutes of Health Study Section on Modeling and Analysis of Biological Systems in 2007-2012. Her service on editorial advisory boards includes the Journal of Computational Chemistry (2004-present), the Journal of Physical Chemistry B (2009-2011), and the Society for Industrial and Applied Mathematics (SIAM) book series on Computational Science and Engineering (2004-2009). She was the editor for Biophysical Journal in 2003-2006.

Dr. Head-Gordon received a B.S. in Chemistry from Case Western Reserve University in 1983, and a Ph.D. in Theoretical Chemistry from Carnegie Mellon University in 1989. She was a postdoctoral member of the technical staff at AT&T Bell Laboratories from 1990-1992.
View Lecture Series Poster



Seminar Series

Related Links

Science at PNNL

Directions

Contacts