Education, Outreach, and Training Track

Sessions represent innovative learning and workforce development applications of TeraGrid resources that help prepare current and future generations to advance scientific discovery. Projects that address the inclusion of under-represented communities have been especially encouraged, including women, minorities, people with disabilities, under-represented institutions, and disciplines that are typically under-represented in HPC.

Sessions address a broad range of EOT activities including:

  • Projects to reform the integration of research and education from K-12 through graduate education addressing multidisciplinary new skills that will help prepare students for competitive global markets of the 21st century, connect concepts across the curriculum inclusive of technology integration, discuss achieved or potential impact, and articulate future plans
  • Creative and successful schemes for engaging today's youth, for partnering with campuses, engaging and retaining under-represented communities, and/or engaging new communities in high-performance computing
  • Successful state-of-the-art strategies for advancing training on high-performance computing, petascale computing, grid computing, and cloud computing
  • Innovative and successful ideas for offering live, synchronous and asynchronous delivery of content to reach people who are separated by time and distance from the instructors and one another.

Schedule of EOT Track Talks:

Block 1: EOT  10 AM – noon (2 hrs – 4 talks)

Computational Science Education: Undergraduate and Graduate Programs (120 minutes)

Steven Gordon and Judith Gardiner. Defining the Competencies for Petascale Computing Education and Training

Download the presentation(PPT)
Abstract: As we move toward the use of Petascale computing, major questions are being raised concerning the readiness of the research community to take advantage of computation at that scale.  A survey of faculty, graduate students, industry, and government labs was completed in spring 2009 (   The 134 respondents provided their rankings of the skills required for undertaking Petascale computing and becoming part of the research community applying those skills to science and engineering research.
Based on those results, we have formulated an initial set of Petascale computing competencies that could become the guide to the creation of the education and training programs in this area.  We initially divided the competencies into two major groups:  basic competencies required for most disciplines and discipline specific competencies.  The competencies touch on seven major categories of skills:

  1. Computer programming
  2. Parallel programming and advanced computer programming
  3. Numerical libraries and algorithms
  4. Debugging, profiling, and optimization
  5. Data management, parallel I/O, and fault tolerance
  6. Scientific visualization
  7. Grid computing

The initial draft competencies are presented along with the feedback received through several virtual meetings with a representative cross-section of researchers.   We will engage the audience in a similar discussion and garner their feedback both at the session and through an on-line survey and forum during and following the Teragrid meeting.  The results will be posted for further discussion along with some initial recommendations for implementing changes in educational curricula and training programs that will promote the appropriate development of a cadre of Petascale capable researchers.

Mary Ann Leung. Two Decades of Impact on the Nation’s High Performance Computing Research: DOE Computational Science Graduate Fellowship Program

Abstract: This talk will provide an overview of the Department of Energy Computational Science Graduate Fellowship (DOE CSGF) program and the impact it has had on the nation’s high performance computing (HPC) research.  The program benefits, requirements, application process, and unique approach to supporting Ph.D. students will be highlighted.  Discussion will also include the approach to building a community of computational science leaders through practical work at DOE laboratories, an interdisciplinary program of study, and a variety of career, professional, and leadership development activities.  Case studies from fellows and alumni will be utilized to illustrate immediate and long term results.  Fellow and alumni stories will exemplify how the program helps students migrate from workstation computing to supercomputing, provides unique opportunities for independent research and collaboration with world renowned experts, and how the program has opened opportunities to people from under-represented communities.

Linda Akli, Jerry Perez and Eduardo Socolovsky. Broadening Participation in the Southeast
Abstract: This is a proposal for a panel discussion moderated by SURA on creative strategies for broadening participation in computational science, high performance computing, and use of campus, regional, and national cyberinfrastructure. The panelists include: Eduardo Socolovsky, Bioinformatics and Computational Science Coordinator for NSU’s Center for Biotechnology and Biomedical Sciences (CBBS); Jerry Perez, Research Associate at TTU’s High Performance Computing Center; and Kelly Robinson, Software Systems Lead at GSU’s academic research computing;. Each of the panelists and their institutions represent three different campus environments, research programs, and approaches to broadening participation; yet have successfully collaborated through SURA IT Programs to support their research and education goals.

Charles Peck, Andrew Fitz Gibbon, Paul Gray, David Joiner, Tom Murphy, Henry Neeman, Skylar Thompson and R. M. Panoff. Teaching High Performance Computing to Undergraduate Faculty and Undergraduate Students
Abstract: A growing proportion of Science, Technology, Engineering \& Mathematics (STEM) research is increasingly dependent on cyberinfrastructure (CI).  CI has experienced rapid progress in enabling technologies -- hardware, storage, networking, middleware, tools, libraries --but much slower improvements in workforce development.  Currently, CI consumers tend to lag substantially behind CI capabilities.  This paper describes a series of linked efforts to address the gap between the workforce and the technology.


Block 1:  EOT  10 AM – noon (2 hrs – 4 talks)
Technology Innovations for Education and Training (60 minutes)
Renato Figueiredo, Arjun Prakash and David Wolinsky. Virtual Appliances for Training and Education in FutureGrid
Abstract: FutureGrid is an experimental computational testbed slated to become part of the TeraGrid infrastructure. FutureGrid provides unique capabilities that enable researchers to deploy customized environments for their experiments in grid and cloud computing. A key enabling technology for this is virtualization and the provisioning of Infrastructure-as-a-Service (IaaS) through cloud computing middleware. FutureGrid's training and education activities leverage these capabilities to create self-contained, flexible, plug-and-play educational "virtual appliances" that serve as a basis for the creation of hands-on educational modules that are easy to share, reproduce and instantiate - on FutureGrid resources, cloud resources, or a user's local resources.
FutureGrid education, training and broader outreach activities include the dissemination of curricular materials on the use of FutureGrid, pre-packaged FutureGrid virtual machines configured for particular course modules, and educational modules based on virtual appliances that focus on education in networking, parallel computing, virtualization and distributed computing. A goal of FutureGrid is to advance education and training in distributed computing at academic institutions with less diverse computational resources, through the development of instructional resources that include preconfigured environments that provide students with sandboxed virtual clusters. Such resources will also provide academic institutions with an opportunity to easily experiment with cloud technology to see if such technology can enhance their campus resources.
FutureGrid educational virtual appliances are designed to be flexible in how and where they are deployed, to provide self-contained and reproducible environments, and to be simple to use. To this end, they support multiple virtualization technologies, including commercial and open-source virtual machine monitors (VMware, VirtualBox, KVM, and Xen), allowing them to run on a variety of resources, including user workstations, desktop grids, cloud resources available in the FutureGrid testbed, and commercial cloud resources (e.g. Amazon EC2 and GoGrid). They expose the same functional behavior when instantiated on different physical resources, and are self-configuring. A core component of the appliance that provides self-configuration and functionally equivalent deployments is its virtual private network – GroupVPN. GroupVPN is a peer-to-peer, self-configuring IP overlay that allows users to deploy independent virtual appliance clusters, automatically creating local- and wide-area virtual clusters from configuration data that is generated through a group Web front-end.
At the conference, the presentation of this abstract will describe the design principles, architecture, and deployment experiences with the virtual appliance technologies used in education and training in FutureGrid. This includes an overview of the GroupVPN virtual network, and the process of configuring and deploying virtual appliance clusters. Examples will highlight three specific educational virtual appliances that cover different middleware stacks used actively in clusters, Grids and clouds: Condor, MPI and Hadoop. The presentation will describe examples of using these appliances in hands-on educational activities, including a TeraGrid introduction to MPI module, with a brief hands-on demonstration using the presenter’s laptop or a video.

Jill Gemmill, Jim Bottum, Peter Cummings, Tom Finholt, Igor Jouline, Stephen Lanier, Ken Lewis, Nicholas Panasik and Jack Wells. A Desktop to Teragrid Ecosystem

Download the presentation(PDF)
Abstract: While the TeraGrid and other High Performance Computing efforts have continued to increase the total computation and storage capacity, the number of active users has not grown at the same pace. Instead, users are significantly increasing their use of local HPC resources. There is a gap in the creation and utilization of the lower tiers of the Branscomb pyramid [1] that, in the long term, is a threat to the health of the scientific computing ecosystem. Our program is a collaboration between the South Carolina and Tennessee EPSCoR jurisdictions, focusing on cyberinfrastructure requirements for Advanced Materials and Systems Biology science domains.  We aim to strengthen the connections at different levels of the pyramid, facilitate the in-migration of new communities into the cyberinfrastructure ecosystem, and facilitate easy movement up and down the pyramid of computational capabilities within disciplines.
Cyberinfrastructure is more than just connecting people via advanced networks and sophisticated applications; CI is about engaging participants in the generation of knowledge and creating opportunities for participants to share expertise, tools, and facilities in powerful ways that have the potential to significantly advance discovery.  An essential enabler of this vision is an ability to coordinate the use of public sector information technology across scales, from desktop to TeraGrid.  We implement a model for that coordination, building a seamless application development and deployment, technical support, training, and personal communications bridge from desktop to TeraGrid.
NSF award EPS-0919440, South Carolina Research Authority, James R. Bottum, linked to EPS-0919436 (University of Tennessee Knoxville)

Dan Fraser, Kay Hunt, Scott Lathrop, Roger Moye, James Barr von Oehsen, Ruth Pordes and Jeff Pummill. Cyberinfrastructure Campus Bridging with Campus Champions
Abstract: We will present a panel on the vision and program of work for Cyberinfrastructure Campus Champions included in a recent proposal to the CI-Team solicitation of NSF. At TeraGrid ’10 we will ask the attendees for input and needs in these areas. As background we will give status on current activities and plans with the attendees and others on short term steps towards realizing these ideas.
Our vision is that faculty, IT professionals, and students on every campus should be able to easily provide and/or utilize cyberinfrastructure (CI) resources available locally, regionally, nationally and internationally as needed to advance science and engineering research and education. We propose to leverage the successful Open Science Grid (OSG) Campus Grid and Sites and the TeraGrid Campus Champions to form Cyberinfrastructure Campus Champions (CI-CC) working together towards the vision. The activity is proposed within the current framework of OSG and TeraGrid projects as a way to make progress in collaboration in this important aspect of bridging between campus cyberinfrastructure and national cyberinfrastructure. Additionally, we propose to collaborate with all other interested CI providers (EDUCAUSE, Internet2, DOE, DOD, etc.) to ensure campus researchers, educators and students are well informed of the resources and opportunities available to them to utilize cyberinfrastructure, and to avoid duplication of effort and to promote cooperation among CI providers and campuses.
As the map indicates, this combined effort will already have made tremendous strides in engaging diverse campuses across the nation in working to achieve the vision. The Champions will be well informed and trained on local, regional and national CI resources and services, with direct access to regional and national CI providers, to be able to assist local users in bridging from their desktop to local and national CI resources in support of computational science research and education.  In the panel we will discuss the three goals of the proposed program of work:
•    Empowering the Campus Champions Virtual Organization by providing the Campus Champion Virtual Organization (VO) with access to resources and people with expertise on CI resources and services. The CI-CC Program will engage additional cyberinfrastructure providers in addition to OSG and TeraGrid, to provide campuses and their users with information and access to the full range of CI resources and services offered both locally and remotely. CI-CC will immerse Champions and CI Providers in multi-media (voice, video, chat, phone) collaborative environments for sharing CI technical information, for encouraging discussion of challenges, and for quickly providing solutions.
•    Providing Solutions for Campus Bridging to CI Resources by identifying and filling gaps in documentation and training materials to empower Champions and campus users to build, sustain, and gain access to CI resources from their desktop. CI-CC will share information on setting up effective environments for utilizing CI resources from desktops to campus, regional and national resources. CI-CC will provide guides for developing and improving campus-wide CI plans.
•    Engaging New and Non-Traditional Campuses and Users to significantly increase the number of researchers, faculty and students involved in learning and applying CI resources to advance science, technology, engineering and mathematics (STEM) research and education. The effort will involve administrators, researchers, faculty, postdocs, graduate and undergraduate students in all STEM fields, with an emphasis on under-represented communities.
Experience has demonstrated that numerous activities, strategies and pathways are needed to engage and support broader participation in campus level cyberinfrastructure and regional and national scale efforts that include grid computing, terascale and petascale science, engineering, and scholarship among researchers and educators.
The Cyberinfrastructure Campus Champions (CI-CC) Program will serve as a source of:
• Local, regional and national CI information and expertise on the campus.
• Start-up accounts to quickly get researchers and educators building and using CI resources.
• Activities of the resources and services of CI Providers for building and using CI.
• Campus high-performance and high-throughput computing needs and requirements.
• CI programs and projects to educate and train in the provision and use of CI.
A CI Campus Champion may engage in a number of the following activities:
• Providing information about CI Provider resources and services to campus researchers and educators.
• Assist faculty and students to understand possibilities to build, run and use a local distributed CI.
• Assist the campus to provide bridges and gateways that support sharing across multiple sites. 
• Participate in CI Provider activities - Users group, Campus Grids activity, Site Administrators forum.
• Be an ombudsperson, on behalf of the campus to CI Providers.
• Host training, communication and awareness sessions for campus personnel about CI Providers.
In summary, we believe the proposed new Cyberinfrastructure Campus Champion Program will help to raise awareness and access to a new range of resources and services in direct support of research and education on campuses across the country.  The program will also provide a strong foundation for collecting community needs and requirements for further improvements for local, regional and national cyberinfrastructure to optimize innovation and discovery among all science and engineering communities.   This panel will introduce a plan for forming this new Program and will utilize the forum to gather feedback from the community to improve upon the strategy and plans, to encourage participation by additional campuses and CI Providers, and to begin the process of forming a strong and vibrant Virtual Organization.

Alex Ramirez, Diane Baxter and Scott Lathrop. A Vision for Cyberlearning and Workforce Development

Abstract:  The National Science Foundation (NSF) Office of Cyberinfrastructure established the Task Force on Cyberlearning and Workforce Development (LWD) to develop a vision and framework for cyberlearning and workforce development to inform the NSF Cyberinfrastructure Framework for 21st Century Science and Engineering (CRF21).  The Task Force is seeking community input to inform its deliberations and report.  Following a brief introduction, the panelist will open the discussion to hear contributions and concerns from the TeraGrid community.  The charge to the Task Force is as follows:
Charge: The workforce for the 21st century must be cyberinfrastructure savvy if it is to be competitive in the international marketplace. Education is no longer K-12, but rather a lifelong endeavor effecting not only future scientist and engineers but also the general citizenry. Cyberinfrastructure serves a dual role in learning and workforce development. First, our next generation of scientists and engineers must be prepared to incorporate the tools of cyberinfrastructure within the context of interdisciplinary research, which requires learning new methods to observe, acquire, manipulate, and store data. Second, the general population must be effectively trained; individuals who experience opportunities to work with and learn through networked environments learn new ways of doing old things or new ways of doing new things, both essential in an increasingly competitive world. The charge to the group comes directly from Chapter 5 of the Cyberinfrastructure Vision for 21st Century Discovery that identified the following goals in the area of workforce development:
•    Foster the broad deployment and utilization of CI-enabled learning and research environments
•    Support the development of new skills and professions needed for full realization of CI-enabled opportunities;
•    Promote broad participation of underserved groups, communities and institutions, both as creators and users of CI;
•    Stimulate new developments and continual improvements of CI-enabled learning and research environments;
•    Facilitate CI-enabled lifelong learning opportunities ranging from the enhancement of public understanding of science to meeting the needs of the workforce seeking continuing professional development;
•    Support programs that encourage faculty who exemplify the role of teacher-scholars through outstanding research, excellent education and the integration of education and research in computational science and computational science curriculum development;
•    Support the development of programs that connect K-12 students and educators with the types of computational thinking and computational tools that are being facilitated by cyberinfrastructure.

Wednesday – Block 2:  EOT   2:30-4:00 – (90 min – 3 talks)

Working with Pre-College Students and Teachers (90 minutes)
Steven Brandt, Chirag Dekate, Phllip LeBlanc and Thomas Sterling. Beowulf Bootcamp: Teaching Local High Schools about HPC
Abstract: The Beowulf Bootcamp is an initiative designed to raise the awareness of and interest in high performance computing in the high schools the area of Baton Rouge, Louisiana. The goal is to familiarize students with all aspects of a supercomputer, giving them hands-on experience with touching or assembling hardware components.  No less significant to the High Performance Computing adventure was an understanding of the software. Students not only installed the operating system and ran benchmarks, but they learned how to program. Our goal for the programming section was ambitious but focused: we sought to give the students a basic understanding of MPI.

Bonnie Bracey Sutton and Vic Sutton. Not Stuck in the Shallow End but Riding the Wave of Opportunity in Teragrid-Inspired Outreach
Abstract: Rather than have teachers, professors and the learning community stuck in the shallow end of computing, those involved in education, outreach and training (EOT) for the Teragrid network have encouraged and supported an initiative called 'Teacher Day' as an outreach to K-12 and pre-service teachers and others who are  influential in the learning community. This paper provides a brief overview of some of the main initiatives underway, and the supporting resources available.

Edee Wiziecki, R. Jay Mashl, Bernard A'cs, Michael Evans and Jeffrey Moore. Computation and Visualization for High School Chemistry Educators and Students at the Institute for Chemistry Literacy Through Computational Science (ICLCS)
Abstract: Improving students’ understanding of chemistry through computation and visualization is a central goal of the ICLCS (, a 5-year NSF Math Science Partnership grant. During three consecutive summer workshops and three academic year courses, two cohorts of 120 rural Illinois high-school chemistry instructors are immersed in computational thinking and training with computational tools. The ICLCS Fellows use materials that they have developed in collaboration with researchers and peers to enhance their teaching. In the first two years of data, we find that the students whose instructors used computational science resources in the classroom showed significantly higher gains than those students whose instructors did not based on a standardized achievement test from the American Chemical Society (ACS). (Results from Year 3 are currently being evaluated.)
One of the computational tools used by ICLCS Fellows is WebMO, a web-based interface to quantum chemistry software, such as Gaussian and GAMESS that empowers users to construct and visualize molecules and to obtain various properties. In preparation for providing computational resources to a vast number of student users, we have developed a scalable, high-performance computing prototype configuration to host a modified version of WebMO. The primary advantage of the web interface is that inputs are generated and output files are parsed automatically, helping to hide the complexities of handling data, thereby enabling the users to focus on the chemistry content and enhancing the learning experience.
As part of the ICLCS, we are also impacting undergraduate chemistry courses at Illinois by (1) infusing general chemist content with virtual laboratories that are replacing video presentations with interactive computational simulation and visualization activities; (2) infusing computational tools and methods into other first and second year chemistry courses; and (3) engaging organic chemistry students in online activities designed to show 3-D representations of complex organic molecules and processes. We believe that institutional change in higher education may increase the likelihood that students who used computational tools in high school will use them again in their further studies. It is estimated that over 1,900 high-school students have been introduced to WebMO through ICLCS, and last year we helped to introduce WebMO to nearly 900 unique undergraduates across three organic chemistry courses with great success. Thus, WebMO is a useful tool for investigating the nature of chemistry at different educational levels and for promoting the sustainability of computational chemistry education.

Contact EOT Co-chairs Pallavi Ishwad (PSC) and Diane Baxter (SDSC).