SC09 Education Program Session Abstracts


Jump to:
Panels | Computational | Biology | Chemistry | Engineering | Math | Physics | Parallel


Panel Session Abstracts

Teach Parallel Panel

Charlie Peck (Earlham College), Tom Murphy (Contra Costa College), Paul Steinberg (Intel)

Tuesday, 5:30 - 6:30, Room C124

Time is no longer a best performance programming tool, by just waiting 18 months for a doubling of performance. The discussion initiated by last year's panel led to the conclusion that parallelism must seamlessly be woven throughout the CS curriculum. This year provides a diabolic panel focused on details; the six panelists will spend 5 minutes each describing what needs to be done, what they have done, and what they plan to do. The second half will be another near free-for-all with the room and the panelists further discussing these hard to implement simply stated goals.

Computational Thinking Session Abstracts

Introduction to Computational Thinking

Bob Panoff (Shodor Education Foundation), Susan Ragan (Maryland Virtual High School of Science and Mathematics)

Saturday, 10:30-11:30 AM, Room B117

Computational Thinking requires quantitative reasoning, analogical thinking, and multi-scale modeling. In this session, we will investigate whether the answers we collect from a variety of sources are correct. We will also see how viewing information in different ways leads to new interpretations. Designed for those who wish to learn how to lead a workshop in computational thinking as well as those who are novices to computational thinking - K12 thru college.

Other People's Models

Bob Panoff (Shodor Education Foundation), Susan Ragan (Maryland Virtual High School of Science and Mathematics)

Saturday, 1:30-3:00 PM, Room B117

Pre-built simulations are valuable resources for learning. In this session, we will demonstrate a variety of web-based simulations that are widely used in K-16 science and math classrooms. We will model an inquiry-based approach for using these tools. Designed for those who wish to learn how to lead a workshop in computational thinking as well as those who are novices to computational thinking - K12 thru college.

Introduction to Systems Thinking

Bob Panoff (Shodor Education Foundation), Susan Ragan (Maryland Virtual High School of Science and Mathematics)

Saturday, 3:30-5:00 PM, Room B117

Systems based modeling is used to demonstrate aggregate change over time. Using Vensim, we will show examples of models illustrating concepts such as equilibrium, exponential growth, interdependent cycles, and feedback. We will investigate the mathematics behind the model using a problem involving compound interest. Designed for those who wish to learn how to lead a workshop in computational thinking as well as those who are novices to computational thinking - K12 thru college.

Advanced Modeling with Vensim

Bob Panoff (Shodor Education Foundation), Steve Gordon (Ohio Supercomputer Center)

Sunday, 10:30-11:30 AM, Room C124

This session will show participants how to apply various analysis tools, modular model building, and scenario tools to build and analyze more sophisticated models. We will also review several techniques for sharing model components among students and as Java applets.

Introduction to Agent Modeling

Bob Panoff (Shodor Education Foundation), Susan Ragan (Maryland Virtual High School of Science and Mathematics)

Sunday, 1:30-3:00 PM, Room C122

Agent-based modeling is used to model situations in which individual objects follow certain rules. After putting those rules into action, we can generalize the behavior of the situation. Using AgentSheets, we will create a model representing the spread of a communicable disease. Designed for those who wish to learn how to lead a workshop in computational thinking as well as those who are novices to computational thinking - K12 thru college.

Computational Thinking with Excel

Bob Panoff (Shodor Education Foundation), Susan Ragan (Maryland Virtual High School of Science and Mathematics)

Sunday, 3:30-5:00 PM, Room C122

The iterative capability of Excel coupled with its Developer’s Toolkit can be used to create both systems-based and agent-based models. Models illustrating diffusion and Punnett Squares will be built to demonstrate the modeling capability of Excel. Designed for those who wish to learn how to lead a workshop in computational thinking as well as those who are novices to computational thinking - K12 thru college.

Finding the Parallel In Nature

Bob Panoff (Shodor Education Foundation), Susan Ragan (Maryland Virtual High School of Science and Mathematics)

Monday, 1:30-3:00 PM, Room B117

This session will concentrate on examples of parallel processing in the natural world and will focus on how to think about these examples computationally. Participants will learn how to apply their modeling skills to a variety of problems.

Computational Biology Session Abstracts

Computational Thinking for Molecular Biology and Sequence Analysis

Ananth Kalyanaraman (Washington State University), Jeff Krause (Shodor Education Foundation)

Saturday, 10:30am-12:00pm, Room C122

The session will explore and describe ways to interpret and understand the computation encoded within the molecular biological systems of nature. Several examples including (but not limited to) central dogma and RNA secondary structure will be covered. A hands-on experiment will help understand the algorithmic modeling for sequence alignment which is a pervasive operation in sequence analysis. The session will also attempt to highlight some of the basic differences in the way computer scientists and biologists think and operate, in order to facilitate interdisciplinary collaboration in computational biology.

Software: NCBI bl2seq web tool
(http://blast.ncbi.nlm.nih.gov/Blast.cgi?PAGE_TYPE=BlastSearch&PROG_DEF=blastn&BLAST_PROG_DEF=megaBlast&BLAST_SPEC=blast2seq

Transcriptomics: From Small-scale to Large-scale

Ananth Kalyanaraman (Washington State University)

Saturday, 3:30pm-5:00pm, Room B119

The session will serve as an introduction to computational methods used for transcriptomics studies such as cDNA transcript assembly, gene discovery and structure prediction, detection of alternative splicing, and Single Nucleotide Polymorphism discovery. Algorithms and software (serial and HPC) suited for small to large-scale studies will be presented. Special emphasis will be laid out on the impact of next-generation sequencing technologies (e.g., 454, Illumina/Solexa). The participants will use the CAP3 program for hands-on exercises modeled using real next-generation sequencing data.

Software:
CAP3 (http://seq.cs.iastate.edu/download.html)

Molecular Evolution and Phylogeny

Ananth Kalyanaraman (Washington State University), Jeff Krause (Shodor Education Foundation)

Sunday, 1:30pm-3:00pm, Room C123

The session will serve as an introduction to the problem of molecular evolution and phylogenetic tree reconstruction, the underlying computational complexity, and the different approaches (e.g., neighbor-joining, maximum parsimony and maximum likelihood). A simple hands-on exercise will use the Jukes-Cantor model to illustrate the idea of Markov-process models, to see how sequences diverge over generations, and to illustrate how the process of phylogenetic reconstruction is limited by evolutionary distance and model assumptions.

Software:
AgentSheets (http://www.agentsheets.com/), Excel(spreadsheet), Mega 4 (http://www.megasoftware.net/)

Biochemical Kinetics and Regulation

Ananth Kalyanaraman (Washington State University), Jeff Krause (Shodor Education Foundation)

Sunday, 3:30pm-5:00pm, Room C117

This session will cover basic principles of simulation of biochemical kinetics using agent-based and system dynamics modeling. The emergence of complexity and control in these systems will be introduced and explored through models of coupling and regulation of reactions. Methods for graphical representation and analysis will also be illustrated.

Software:
AgentSheets (http://www.agentsheets.com/), NetLogo (http://ccl.northwestern.edu/netlogo/), Vensim PLE (http://www.vensim.com/freedownload.html), Scilab (http://www.scilab.org/platform/)

Modeling for Systems Biology

Jeff Krause (Shodor Education Foundation)

Monday, 10:30am-11:30am, Room B119

This session will extend on ideas introduced in the session on biochemical kinetics and regulation to address the challenges of building models of molecular interactions and reactions that account for cellular and higher order phenomena. Resources that support these efforts will be described as well as model repositories, standards and tools.

Software:
CellDesigner (http://www.celldesigner.org/), Scilab (http://www.scilab.org/platform/)

Homology Modeling I

Masa Watanabe (University of California, Merced)

Monday, 1:30pm-3:00pm, Room C122

This session will introduce homology (comparative) modeling, which is used to predict the 3D-structure of a unknown protein based on the known structure of a similar protein. In Part I session, the general idea of homology modeling will be introduced. In addition, a prediction the three-dimensional structure of a given protein sequence based on an alignment to one or more known protein structures will be demonstrated.

Software:
Swiss-Model Server (http://swissmodel.expasy.org/) Swiss-Pdb Viewre (http://spdbv.vital-it.ch/)

Homology Modeling II

Masa Watanabe (University of California, Merced)

Monday, 3:30pm-5:00pm, Room C122

This session will continue to introduce homology (comparative) modeling. The main goal of this session is to provide you with a basic framework with how to create and validate a homology model. Also, in this tutorial, molecular dynamics simulation method will be introduced to make detail structural and dynamical analysis on the validated homology model.

Software:
Swiss-Model Server (http://swissmodel.expasy.org/), VMD(http://www.ks.uiuc.edu/Research/vmd/), OpenMM Zephyr (https://simtk.org/home/zephyr)

Introduction to R for Biological Computing

Jeff Krause (Shodor Education Foundation)

Tuesday, 1:30pm-3:00pm, Room B117

This session will introduce the R statistical programming language and some of it’s many uses in biological modeling and analysis. The power and flexibility of this high-level scripting language, with its broad range of statistical functions, and graphing capabilities along with the ease of authoring new packages including datasets and functions, have lead to widespread use of this tool across the biological sciences. This session will introduce some basic functionality and useful collections of resources for this open-source tool.

Computational Chemistry Session Abstracts

Teaching Chemistry Using CCDC Database

Clyde Metz (College of Charleston), Shawn Sendlinger (North Carolina Central University)

Saturday, 1:30pm - 3:00pm, Room B118

The Cambridge Crystallographic Data Centre (CCDC) provides a free 500-structure interactive, online database for use in educational applications. Visualization of the structures utilizes the free version of the popular Mercury program. A variety of exercises will be provided that show how this database and visualization software can be used to assist in teaching chemical concepts such as VSEPR, aromaticity, and stereochemistry.

Software:
Mercury http://www.ccdc.cam.ac.uk/free_services/mercury
Teaching Modules http://www.ccdc.cam.ac.uk/free_services/teaching/modules

Properties of Molecules

Clyde Metz (College of Charleston), Shawn Sendlinger (North Carolina Central University)

Saturday, 3:30pm - 5:00pm, Room B118

This largely hands-on session will demonstrate the use of two free software packages that can be used to draw molecular structures of publication quality, visualize three-dimensional molecular structures, calculate stable molecular structures using molecular mechanics, and predict various molecular properties.

Software:
ACD/ChemSketch Freeware http://freechemsketch.com/
KnowItAll http://www.knowitall.com/academics/welcome.asp

Chemical Kinetics

Clyde Metz (College of Charleston), Shawn Sendlinger (North Carolina Central University)

Sunday, 10:30am - 11:30am, Room B117

A basic review of first-order chemical kinetics will be followed with a hands-on session using Excel (a spreadsheet) and Vensim PLE (a graphical interface differential equation solver or system dynamics software) to study the concentration-time relationship for a kinetics system.

Software:
Vensim PLE http://www.vensim.com/freedownload.html
Spreadsheet (Excel)

Basic Molecular Modeling

Shawn Sendlinger (North Carolina Central University), Clyde Metz (College of Charleston)

Sunday, 1:30pm - 3:00pm, Room B117

This session will serve as an introduction to molecular modeling and will discuss reasons why it should be included in the chemistry curriculum. The hands-on portion will introduce participants to the WebMO interface used to construct molecules, perform simple calculations, and view.

Software:
WebMO accounts at Earlham and/or NCSA

National Science Digital Library Using The Computational Science Education Reference Desk

Shawn Sendlinger (North Carolina Central University), and Clyde Metz (College of Charleston)

Monday, 3:30pm - 5:00pm, Room B117

This session will serve as an introduction to molecular modeling and will discuss reasons why it should be included in the chemistry curriculum. The hands-on portion will introduce participants to the WebMO interface used to construct molecules, perform simple calculations, and view.

Software:
WebMO accounts at Earlham and/or NCSA

Intermediate Molecular Modeling

Shawn Sendlinger (North Carolina Central University), and Clyde Metz (College of Charleston)

Monday, 3:30pm - 5:00pm, Room B119

This largely hands-on session will introduce the structure and use of the Computational Science Education Reference Desk (CSERD) which serves as a pathway to the NSDL. Participants will be given the opportunity to submit a review of a software package of interest to them.

Computational Engineering Session Abstracts

Modeling in Engineering using MATLAB

Steve Gordon and Siddharth Samsi (Ohio Supercomputer Center)

Saturday, 10:30am - 12:00pm, Room B118

We will introduce the use of MATLAB for basic understanding of modeling and simulation principles applied to instruction for undergraduate engineers. Participants will work through a simple exercise on traffic using MATLAB and Excel that is used to introduce students to the terminology and purposes of modeling and simulation.

Simulink Basics for Engineering Applications

Ashok Krishnamurthy (Ohio Supercomputer Center), Siddarth Samsi (Ohio Supercomputer Center)

Saturday, 3:30 pm - 5:00 pm, Room C122

Simulink®, from the MathWorks Inc. is modeling software that enables users to create a model that represents the problem and provides mechanisms to examine the model as it runs. This course will present the basics of the Simulink system and introduce participants to the most commonly used tools in the Simulink libraries. Participants will have the opportunity to build and simulate a model and learn tools that can they can apply to their problem domain.

Introduction to Computational Fluid Dynamics

James Giuliani (Ohio Supercomputer Center)

Sunday, 10:30am to 11:30am, Room B118

An introduction will be given to the field of Computational Fluid Dynamics (CFD) and the types of physical systems that can be modeled. Participants will use several Java applets to solve typical computational fluid dynamics problems and learn about the type of numerical results that can be obtained. The applications will be discussed in light of the basic equations governing fluid dynamics and the nature of the simplifications embedded in the applets. The process of building, solving and analyzing a CFD problem will be examined to highlight how the physical system is discretized into numerical form so that the governing equations can be applied and a solution obtained.

Advanced Computational Fluid Dynamics

James Giuliani (Ohio Supercomputer Center)

Sunday, 1:30pm to 3:00pm, Room B118

Participants will examine more advanced capabilities and characteristics of Computational Fluid Dynamics (CFD) by solving several example problems with the OpenFOAM (Open Field Operation and Manipulation) CFD Toolbox. Access to the OpenFOAM software running on a remote supercomputer will be provided through the Ralph Regula School Computational Fluid Dynamics web portal interface, in addition to direct access to the software. Topics of initial conditions, boundary conditions, and numerical stability will be discussed within the context of the example problems. Information regarding the governing fluid dynamic equations and the basic discretization techniques used to numerically approximate them will also be presented.

Chemical Engineering Thermodynamics Modeling

Michael Paulaitus (Ohio State University)

Sunday, 3:30pm - 5:00pm, Room B118

The free energy plays a central role in thermodynamic analyses of chemical engineering unit operations. This central role is evident in undergraduate chemical engineering thermodynamics courses where emphasis is placed on applying analytical equations of state to calculate free energies. Indeed, the advent of scientific computing enabled chemical engineers to develop and implement more complex, accurate, and computationally robust analytical equations of state with broad applicability, but in large part for systems relevant to the petrochemical industries. More recent chemical engineering applications to biological systems have led to an interest in aqueous solution thermodynamics, and free energy calculations for complex biomolecular and supramolecular systems, for which equations of state do not exist that can incorporate the molecular details needed to describe the essential elements of biological structure and function. Thus, chemical engineers have turned to calculating free energies directly from molecular simulations. Chemical engineering education has also begun to introduce molecular simulation-based free energy calculations in core thermodynamic courses as a tool for practical calculations, as well as a tool for demonstrating thermodynamic concepts, most notably entropy. In this presentation, we take the approach that molecular simulation data are available, much like experimental data on actual systems of chemical engineering interest might be available, and ask students calculate free energies from these data using various numerical/computational techniques. Modules are developed to familiarize students with these different techniques, and ultimately to apply them to solve several problems involving self-assembly in aqueous solutions.

Mechanical Engineering Applications

Dan Warner (Clemson University)

Monday, 10:30 - 11:30, Room C124

This session will showcase material from a new course at Clemson University that has been developed for sophomore and junior students in Mechanical Engineering. This workshop will review the development of both programming skills using MATLAB, as well the mathematical and physical knowledge to prepare the students to analyze an indeterminate planar truss.

The students will have completed an elementary course is statics and can set up and solve a simple determinate truss by hand. They have also had a brief introduction to MATLAB.

The 6 major steps along the path to analyzing an indeterminate truss are:
1. Develop a computational approach to specifying and displaying a truss.
2. Develop the mathematical background in matrix algebra and the numerical solution of linear systems of equations.
3. Develop the MATLAB codes for constructing and solving the linear system based on the elementary summation of forces.
4. Develop the formulation of potential energy based on virtual displacements.
5. Develop the mathematical background in matrix algebra for minimizing a quadratic function of several variables.
6. Develop the MATLAB code for constructing and analyzing the linear system that minimize the potential energy.

The workshop participants will be provided with MATLAB codes and handouts that reflect each of the 6 steps. The codes will be executable, but will contain annotations indicating which portions the students would normally be required to develop in their own.

Bioengineering with COMSOL Multiphysics

Richard Hart (Ohio State University)

Monday, 1:30pm - 3:30pm, Room B118

The COMSOL (COMSOL, Inc., Burlington, MA) multi-physics finite element package is particularly well suited for solving problems in biomedical engineering. Not only can biomechanics problems be formulated and solved (including solid and fluid mechanics), but other phenomena, including bioelectricity, biotransport, bioimaging, bioacoustics, bioheat, etc. can be modeled in arbitrary combinations. COMSOL and tutorial problems (many of which are available to download from the COMSOL website) involving total hip replacements, vascular stents, pacemaker electrodes, and others will be demonstrated and available for modification and solution.

Parallel MATLAB

Siddharth Samsi and Ashok Krishnamurthy (Ohio Supercomputer Center)

Monday, 3:30 pm - 5:00 pm, Room B118

This course will present methodologies on exploiting parallelism in MATLAB® programs using the Parallel Computing ToolboxTM (PCT) and the MATLAB Distributed Computing ServerTM (MDCS). Participants will be introduced to concepts of task and data parallel programming. The course will specifically cover various parallel programming constructs offered by the PCT including parallel for-loops, distributed arrays and message passing. Example code will be presented and made available to participants for modification and/or incorporation into their own applications. Participants will also have the opportunity to develop a series of real world signal and image processing examples intended to reinforce the concepts of task parallel and data parallel programming with MATLAB®.

Advanced Simulink for Modeling and Simulating Systems

Ashok Krishnamurthy (Ohio Supercomputer Center), Siddarth Samsi (Ohio Supercomputer Center)

Tuesday, 1:30 PM - 3:30pm, Room B118

Simulink, an add-on product to MATLAB, provides an interactive, graphical environment for modeling continuous and discrete time systems. Besides using the provided block libraries, a user can create custom processing blocks in C or MATLAB, create subsystems, and develop libraries of blocks. This module will build on the Introduction to Simulink module, and introduce creating your own custom blocks and libraries. We will also discuss using code generation using the Real-time Workshop. Examples from audio signal processing and automotive applications, and several hands-on laboratories will be used to illustrate the concepts. Prerequisites: Working knowledge of MATLAB, Introductory Simulink

Parallel MATLAB II: Creating a Parallel Application

Ashok Krishnamurthy and Siddharth Samsi (Ohio Supercomputer Center)

Tuesday, 3:30pm - 5:00pm, Room B118

This course will utilize concepts discussed in the Parallel MATLAB tutorial and apply them to a selected algorithm. Participants will be walked through the process of taking a serial MATLAB algorithm and using the Parallel Computing Toolbox TM (PCT) to first develop and debug a parallel version of the algorithm and then scale the algorithm to larger problem sizes using the MATLAB Distributed Computing Server TM (MDCS). Participants will have the opportunity to analyze the serial algorithm and create a parallel version of the algorithm using parallel for-loops as well as through the use of distributed arrays through the use of the PCT and the MDCS.

Computational Math Session Abstracts

Computational and Experimental Mathematics: C & E Math, an Introduction

Dan Warner and Neil Calkin (Clemson University)

Saturday, 1:30 - 3:00, Room C122

The Twentieth Century witnessed a tectonic shift in our understanding and practice of mathematics.

During the last 60 years the computer has played an increasingly important role in the development of mathematics, ranging from its key but controversial use in the proof of the Four Color Theorem; to its essential use in the development and promotion of Fractals and Chaos Theory; to its ubiquitous use, particularly in discrete mathematics, as an exploratory tool. This workshop will provide a hands-on introduction to one of the most recent developments in computer tools for mathematics, the SAGE Mathematics Computing Environment. SAGE is a comprehensive package which addresses mathematical topics ranging from rational arithmetic and elementary algebra to some of the most advanced research topics in Number Theory and Discrete Mathematics. In addition to its wide ranging collection of tools, SAGE is important because it is Open Source, free, and accessible with nothing more than a modern web browser.

This introductory session will focus on using SAGE as a tool for exploring elementary topics in mathematics. The session will be accessible to mathematics teachers at all levels. The topics that will be explored will be restricted to precalculus material, and potential pedagogical applications will be highlighted. However, be prepared to learn some new facts about pi, Fibonacci numbers, and Mandelbrot's set.

Computational and Experimental Mathematics: C & E Math, Intermediate Topics

Dan Warner and Neil Calkin (Clemson University)

Monday, 1:30 - 3:00, Room C124

This workshop will continue the hands-on introduction to the SAGE Mathematics Computing Environment. SAGE is a comprehensive package which addresses mathematical topics ranging from rational arithmetic and elementary algebra to some of the most advanced research topics in Number Theory and Discrete Mathematics. In addition to its wide ranging collection of tools, SAGE is important because it is Open Source, free, and accessible with nothing more than a modern web browser.

This session will focus on using SAGE as a tool for exploring intermediate mathematical topics, particularly topics that could be encountered in Calculus and other undergraduate courses.

The presentation will include material on some of the advanced graphics features, such as phase plane portraits and surface plots, as well as standard problems in symbolic differentiation and integration. In addition the workshop will provide some exploratory problems that may well lead the participants to some new insights.

Computational and Experimental Mathematics: C & E Math, Advanced Topics

Dan Warner and Neil Calkin (Clemson University)

Monday, 3:30 - 5:00, Room B117

This workshop will continue the hands-on introduction to the SAGE Mathematics Computing Environment.

SAGE is a comprehensive package which addresses mathematical topics ranging from rational arithmetic and elementary algebra to some of the most advanced research topics in Number Theory and Discrete Mathematics. In addition to its wide ranging collection of tools, SAGE is important because it is Open Source, free, and accessible with nothing more than a modern web browser.

This session will focus on using SAGE as a tool for exploring advanced topics related to recent mathematical research. The will include an overview of the advanced programs such as GMP and PARI that are included in the SAGE package. The session will also include a presentation on integer relation algorithms and some recent results.

Computational Physics Session Abstracts

Examples of Computation to Use in Physics Courses

Rubin Landau (Oregon State University)

Saturday, 10:30 am - 11:30 am, Room B119

This session assumes participants have some familiarity with WebMO or have attended the Basic Molecular Modeling session. More advanced applications that include spectroscopic calculations (Infrared, Ultraviolet-Visible, Nuclear Magnetic Resonance) will be discussed. The hands-on portion will provide experience in performing and interpreting these calculations, and will also demonstrate how utilizing multiple processors can reduce computational expense.

Software:
WebMO accounts at Earlham and/or NCSA

N-Body I

Dave Joiner (Kean University), Norman Chonacky (Yale University)

Saturday, 1:30 pm - 3:00 pm, Room B119

The complexity of the interaction of objects increases greatly with the number of objects being considered, whether one is considering gravitational, electromagnetic, or inter-atomic forces. Session 1 will focus on the classes of problems for which direct methods are feasible and typical simulations based on these methods that can be used interactively in the classroom. Exercises will include runnable canned simulations, editable Java simulations, and C/C++ simulations parallelized using MPI.

Pathways

Dave Joiner (Kean University), Norman Chonacky (Yale University), Mario Belloni (Davidson College), Richard Gass (University of Cincinnati)

Sunday, 10:30 am - 11:30 am, Room B119

Over the last 3 years, the Physics team for SC Education has been working to define the set of problems in computational physics that exemplify the overlap between the standard physics curriculum, fundamental computational physics skills, and advances in high performance computing. The direction of the physics team will be discussed in an open forum, along with opportunities to work with the physics team and provide feedback on pursuing key problems and identifying existing content for further dissemination.

EJS - Easy Java Simulations

Mario Belloni (Davidson College)

Sunday, 1:30 - 3:00, Room B119

In teaching science, many of the topics taught require mathematics or visualizations that are not familiar to the students in these courses. To address this issue, we have created a set of flexible resources for the teaching of science based on two- and three-dimensional simulations. They are created with Easy Java Simulations, EJS, which is a free and open source authoring tool. Because EJS allows teachers to easily change simulations, existing simulations can be customized to the level and type of science course one is teaching. This workshop will show examples including simple 1- and 2-dimensional motion, population biology, the seasons, and eclipses (solar and lunar). Workshop participants will also have the oppoirtunity to modify and examine existing models in order to gain experience with Easy Java Simulations. The EJS authoring tool and examples of existing models from astronomy, biology, and physics will be distributed on CD and are also available on ComPADRE, an NSDL Pathway.

N-Body II

Dave Joiner (Kean University), Norman Chonacky (Yale University)

Sunday, 3:30 pm - 5:00 pm, Room B119

Continuation of N-Body I, this session will focus on scalability issues and key algorithms that are useful for N-body methods. In particular, we will deal with the Barnes-Hut and Particle-Mesh approximations and investigate the program structures and methods, often not treated in computational courses, typically needed to implement them. We will also examine options for implementing those algorithms on parallel systems and discuss the implications of treating this class of problems either interactively, in a typical class period, or in a longer-term student project. Exercises may include running parallel N-body simulations on remote clusters.

Digital Libraries

Dave Joiner (Kean University), Norman Chonacky (Yale University), Mario Belloni (Davidson College), Richard Gass (University of Cincinnati)

Monday, 10:30 am -11:30 am, Room B118

Many digital library projects in Physics currently exist to catalog and disseminate educational materials that include computation. Among the digital library efforts that will be discussed are ComPadre and the Computational Science Education Reference Desk. The ComPadre digital library focuses on Physics Education, and includes sections on pedagogical articles and information, traditional lesson plan, and computational activities. Through a partnership with the Open Source Physics projects, ComPadre hosts an extensive selection of Physlet, OSP, and Easy Java Simulation models for physics teachers. The Computational Science Education Reference Desk focuses on computational science education, and includes sections on parallel algorithms and on computational physics activities. Other digital library efforts will also be discussed in a panel format with demonstrations.

Introduction to Parallel Computing I

Richard Gass (University of Cincinnati)

Monday, 1:30 pm - 3:30 pm, Room B119

This will be a hands-on introduction to parallel computing with Mathematica. We will focus on problems of interest in physics that are accessible to undergraduates. No previous experience with Mathematica or parallel computing is necessary although have worked through the tutorial “The first Five minutes with Mathematica” would be helpful. Just launch Mathematica on the laptop and it will come up.

Parallel Computing II - Applications

Richard Gass (University of Cincinnati)

Tuesday, 1:30 pm - 3:30 pm, Room C122

A continuation of part one with an emphasis on physical applications. We will also look at an example that is not embarrassingly parallel namely solving PDE’s by domain decomposition. This is a hands-on session.

Computational Physics/Chemistry with ABINIT

Jeff Rufinus (Widener University)

Tuesday, 3:30 pm - 5:00 pm, Room B117

ABINIT is a free, open source code that can be used to do a variety of first principles calculations. The code is based on Density Functional Theory. The physical properties of systems of materials (e.g. molecules, nanostructures, periodic structures) can be studied using the code. The parallel ABINIT code is also available in order to speed up the calculation of larger systems. In this talk, the ABINIT code and its capabilities will be presented. Some examples, ranging from molecules, nanostructures, and periodic structures, will also be given. The emphasis of this presentation is in the hands-on experience, in which each participant will be given the opportunity to use the code. In order to use the code, each participant will be asked to write some simple input files. The input files will be run and the output files will be examined.

Parallel Programming Session Abstracts

What the Heck is Supercomputing?

Charlie Peck (Earlham College)

Saturday, 10:30 - 11:30 AM, Room C123-124

An overall introduction to high performance computing designed for computer scientists and disciplinary scientists. The many forms of parallelism found in modern hardware, and the principle software architectures used with them, distributed and shared memory programming, will be examined. In some ways this talk builds the framework for subsequent talks in the Parallel track; the storage hierarchy, MPI, OpenMP and GPGPU programming will all be placed in the larger context of high performance computing.

Tyranny of the Storage Hierarchy

Charlie Peck (Earlham College)

Sunday, 10:30 - 11:30 AM, Room C122

Modern computer architectures have an increasingly complex and non-uniform memory hierarchy, as the core-count continues to grow this is likely to be exacerbated further. Understanding the technical details of how memory busses operate and how they can be used effectively are key to building efficient parallel applications. This session will include a hands-on lab which will explore the material from an experimental perspective.

Bootable Cluster CD with MPI Examples

Leandro Avila (University of Northern Iowa)

Sunday, 1:30-3:00 PM, Room C124

The Bootable Cluster CD (BCCD) is a tool for teaching parallel and cluster computing in the classroom. Not all institutions have access to cluster resources, but this live CD can be used to turn any lab of Windows, Mac, or Linux computers into a temporary but fully functional cluster, complete with support for MPI. The CD is completely non-destructive and can also be used by students to have the same computing environment at home on their computers. In this session, participant's PCs will be booted up and turned into a cluster, and the MPI parallel programming examples included on the CD will be demonstrated along with a discussion of their use in the curriculum.

Message Passing Interface (MPI) Programming I

Tom Murphy (Contra Costa College)

Sunday, 3:30 - 5:00, Room C124

MPI is the standard most typically used to write a distributed memory program running on parallel architectures such as a cluster. In this session, participants will learn the essential MPI programming framework and the seven most used MPI library calls, which not only allow first programs to be written, but are also the library calls most heavily used in any MPI based program. Participants will use workshop laptops to dissect a working MPI program and use it to write one of their own.

openMP Programming

Tom Murphy (Contra Costa College)

Monday, 10:30 - 11:30, Room C122

openMP is the mechanism most typically used to write a shared memory program to running parallel architectures such as a multicore computer. In this session, participants will learn the most common compiler directives, library calls, and environment variables used to coax a single core program to become thread aware and successfully run on multiple cores. Participants will use workshop laptops to run a series of working and non-working sample openMP programs as a method of their breaking the ice with openMP.

CUDA Programming

Leandro Avila (University of Northern Iowa)

Monday, 1:30pm - 3:00pm, Room C123

Harness the power of modern graphic processing units to improve performance of highly-threaded code. Introduction to CUDA programming for scientific computing. CUDA architecture overview, API introduction and discussion. Why the CUDA architecture is so different from the traditional model, and more importantly, why you should care.

Message Passing Interface (MPI) Programming II

Kay Wanous (Earlham College)

Tuesday, 3:30pm - 5:00pm, Room C122

This is the second session in the MPI (the Message Passing Interface) series and will build upon the first one. Basic knowledge of MPI is suggested. This session will introduce various methods of collective communication and their uses. MPI_Bcast, MPI_Scatter, MPI_Gather, MPI_Reduce, and MPI_Allreduce will be introduced. Lab exercises will be provided.