Future Cyberinfrastructure

min read

© 2008 George B. Adams III
EDUCAUSE Review, vol. 44, no. 1 (January/February 2009): 8–9.

Future Cyberinfrastructure

By George B. Adams III

George B. Adams III is Deputy Director and Associate Director for Programs, Network for Computational Nanotechnology (home of nanoHUB.org), and Director, HUBzero Project, at Purdue University.

Comments on this article can be sent to the author at <[email protected]> and/or can be posted to the web via the link at the bottom of this page.

The following excerpt is based on an interview conducted by Gerry Bayne, EDUCAUSE multimedia producer, at the Coalition for Networked Information (CNI) 2008 Spring Task Force Meeting held in Minneapolis, Minnesota. To listen to the full podcast, go to <http://www.educause.edu/blog/gbayne/CNIPodcastnanoHUBorgFutureCybe/167636>.

Gerry Bayne: nanoHUB.org has developed a suite of tools for use with the Open Science Grid and TeraGrid. Can you describe these computing grids, their accompanying complexities, and how nanoHUB has been able to hide these distractions to let the researcher concentrate on research?

George B. Adams III: This question gets to the heart of what I consider to be the key nanoHUB.org concept: delivering value to the nanotechnology research and education community. TeraGrid and the Open Science Grid are two examples of very large, distributed computational resources that serve the science community. TeraGrid is funded by the National Science Foundation (NSF) and comprises eleven partner sites, including Purdue University, and provides access to high?performance computers, data resources and tools, and high?end experimental facilities around the United States. The Open Science Grid is a consortium of software, service, and resource providers and researchers from universities, national laboratories and computing centers across the United States. Funded by the NSF and the U.S. Department of Energy’s Office of Science, it helps satisfy computing and data management requirements for scientific researchers with a special focus on collaborative science requiring high?throughput computing.

Access to TeraGrid resources is by awards made to individual principal investigators in response to successful proposals reviewed by a national peer review committee. The Open Science Grid brings together formally defined and recognized groups of people for the contracted or opportunistic sharing of the Open Science Grid resources. So, to reduce the time to begin to use grid computing, nanoHUB.org has built a relationship with TeraGrid and the Open Science Grid. The nanoHUB team has been awarded, through peer review, several continuous community allocations on TeraGrid, and the Open Science Grid also provides allocations to Purdue through its partnership. By way of the new concept of community accounts, a researcher or student who registers for a free nanoHUB account has access to grid computing power to run nanoHUB tools served with graphical user interfaces without any further administrative action. This makes access easier. Second, nanoHUB.org is powered by software that helps automate the interaction of the applications with the grid computing resources. We have more work to do to improve the robustness of this automation and to increase its capability to manage more complex jobs, but this again helps the researcher and the student focus on research or education and not on the management of the computation. So, access is easier, and use is easier—two ways that nanoHUB.org delivers value to the research and education community.

Bayne: Can you talk about the fields of focus, including nanoelectronics and nano/bio?

Adams: Sure. nanoHUB is designed to be a science gateway for the entire nanotechnology community. It is a creation of the NSF-funded Network for Computational Nanotechnology (NCN). NCN connects computation and experiment to address research challenges and to develop community simulation codes for nanotechnology. The area of nanoelectronics is particularly well-suited for this because there is a history of development of open-source community simulation codes, such as the electronic circuit code SPICE, and thus a familiarity with the open sharing of programs to model electronic phenomena. Extending this to nanoelectronics has been very fruitful.

Nano/bio is an area of particularly strong, early promise in nanotechnology. The unexplored territory of this scientific discipline is so large that modeling and simulation are particularly valuable in guiding experiment and interpreting surprising results. In fact, one of the most recent additions to nanoHUB is a tool called BioMOCA, the first module of a suite of tools to allow experimentalists to design nanoelectronic sensors for biomolecules. For these reasons, nanoHUB.org initially focused on nanoelectronics, nano/bio, and nanoelectromechanics. As nanoHUB grows, new areas are being supported, including nanophotonics.

Bayne: What sorts of researchers have been using nanoHUB, and are there any underutilized elements of this tool?

Adams: The most utilized tools and informational items are those that serve the community particularly well. nanoHUB users are encouraged to rank and review all content contributed to nanoHUB.org. Robust tools that focus on important questions are heavily used, and polished presentations on important topics are widely viewed.

We track the top-ten tools list with respect to number of users, number of simulations run with that tool, and number of CPU hours consumed. We do see fluctuations in these rankings determined by fluctuating demands in classroom and research usage. Currently, nanoMOS, a two?dimensional simulator for thin?body MOSFET transistors, and Nanosphere Optics Lab are two of the most heavily used tools. Presentations on MOSFET theory and thermal conduction in nanostructure materials are very popular. Materials of lesser quality or interest naturally are not ranked as highly and fall to lower levels of usage.

Bayne: Are some faculty beginning to use nanoHUB in conjunction with their teaching, and are some students beginning to use the suite of tools directly in some of their assignments?

Adams: Yes, faculty are using nanoHUB in conjunction with their teaching. Professor Supriyo Datta, of Purdue University, placed his course "Quantum Transport: Atom to Transistor" on nanoHUB. This Ph.D.-level nanoelectronics course has been viewed by over 6,300 persons since it was posted online on August 7, 2006. Professor Datta taught this course on campus to 60 Purdue students in the classroom. So the impact he can have with his new ideas about engineering nanoelectronic devices increased from 60 students to 6,300.

And yes, simulation tools available on nanoHUB.org are being directly used by students. Professor George Shatz, of Northwestern University’s Department of Chemistry, has all freshman chemistry students use the tool QC-Lab as part of their laboratory program. Professor Philip Wong, of Stanford University's Department of Electrical Engineering, uses nanoHUB simulation tools in his undergraduate class "Introduction to Nanoelectronics and Nanotechnology," to teach about resonant tunneling diodes. One of his students from the 2005 offering of that course went on to use the tool more extensively as a graduate student and published a paper in an IEEE journal in 2007 with Professor Wong.

I believe that simulation tools should be a part of education and a component of technological literacy. The question of how to best use simulation in a classroom and what impact it can have on the learning experience is of great interest to us. To answer that question, we have partnered with Purdue Professors George Bodner (Chemistry Education), David Radcliffe (School of Engineering Education), and Sean Brophy (Aeronautics and Astronautics and the School of Engineering Education). Professor Bodner conducts research to find better ways of conveying information to students. Professor Radcliffe's research interests include the design of creative learning places. Professor Brophy is investigating how he can develop students’ abilities to understand, troubleshoot, and ultimately design complex systems using a combination of analogical reasoning, model construction and evaluation, and simulation. I believe the research that this team is now carrying out will help us understand how to use simulation tools, such as those on nanoHUB.org, effectively for education.

Bayne: What is HUBzero, and how is it related to nanoHUB?

Adams: HUBzero powers the nanoHUB.org website. HUBzero provides middleware that makes it easy to create dynamic science gateways to host interactive computer programs and to connect and support a scalable community of scholars. In addition to middleware for accessing simulation tools and computational resources, HUBzero also provides simple mechanisms for the community to contribute to a website—sharing their content, collaborating, and discussing their research or educational materials. HUBzero software provides a scalable solution that enables the nanoHUB team to support, with a very limited support team, more than 85,000 users today.

Bayne: What are some of the future projects and developments planned for nanoHUB?

Adams: Very soon you'll see a recommendation engine as part of the site, digital object identifiers for simulation tools and versioning, and an incentive system to contribute questions and answers to nanoHUB.org. We're also working on a general optimization engine that can drive the simulation tools toward certain design targets that a user might have. The hub concept itself is being enthusiastically embraced by new communities forming around topic areas, including "Heat Transfer" and the "Manufacture of Pharmaceuticals": thermalHub.org and pharmaHub.org are two examples of a rapidly growing set of sites powered by HUBzero. We believe there will be many more, and we look forward to working with research and education communities to identify and deliver valuable services in support of their work.