Computing in Science and Engineering

Coming to a Digital Library Near You

Do you know about the National Science Digital Library (NSDL)? You should, no matter what your field or occupation within science, technology, engineering, or mathematics (so visit http://nsdl.org/about). Why? Because sometime this autumn a new resource within the NSDL, called UCOMP that's dedicated to computation in physics, will join this digital collection. Editor-in-chief Norman Chonacky describes it all in this installment of his First Word column

Computational Methods with Depth and Flair

Stephen Weppner reviews the third edition of An Introduction to Computer Simulation Methods: Applications to Physical Systems.

High-Performance Computing Education

Many successful efforts are currently addressing the critical shortage of a diverse, well-prepared high-performance computing (HPC) workforce. The guest editors' goal, with this issue, is to stimulate large-scale international discourse to accelerate the adoption of the educational tools, curriculum, and pedagogy that reflect the increasing role of computational methods in science and engineering. Such approaches are necessary to educate a larger and more diverse population of well-skilled, knowledgeable, and innovative people who will significantly advance scientific discovery in all fields of study, now and well into the future.

A Shared, Interinstitutional Undergraduate Minor Program in Computational Science

Nine Ohio higher education institutions combined expertise to implement an interinstitutional, competency-based undergraduate minor curriculum in computational science. The program, which began in Autumn 2007, is part of an National Science Foundation Cyberinfrastructure Team Award and the new, statewide virtual Ralph Regula School of Computational Science. Key program elements include the definition of the program's interdisciplinary competency requirements, sharable instructional modules, and interinstitutional agreements for cross registration. Many courses are offered remotely, so students can access computational science instruction likely unavailable at their home institutions. The program is a model for interdisciplinary, interinstitutional activities that provide the expertise required to advance the application of computational methods to new discoveries in science and engineering.

nanoHUB.org: Advancing Education and Research in Nanotechnology

In 2002, the National Science Foundation established the Network for Computational Nanotechnology (NCN), a network of universities supporting the National Nanotechnology Initiative by bringing computational tools online, making the tools easy to use, and supporting the tools with educational materials. Along the way, NCN created a unique cyberinfrastructure to support its Web site, nanoHUB.org, where researchers, educators, and professionals collaborate, share resources, and solve real nanotechnology problems. In 2007, nanoHUB.org served more than 56,000 users from 172 countries. In this article, the authors share their experiences in developing this cyberinfrastructure and using it, particularly in an educational context.

Applying Computational Science to Education: The Molecular Workbench Paradigm

The Molecular Workbench offers highly interactive molecular dynamics simulations to help students learn difficult scientific concepts. The software demonstrates how scientists can transform research tools into educational tools. Research studies show that students learn better using computational models.

Transforming Scientists through Technical Education: A View from the Trenches

Critical, emerging problems necessitate finding new ways to study the environment. To this end, technology innovators must school researchers in an array of emerging technologies and teach them how to apply such technologies to their work. Two research projects took on this challenge, creating successful workshops and a virtual seminar designed to provide problem-based instruction. Here, project leaders evaluate their efforts and assess their activities, outcomes, and lessons learned.

Transforming Chemistry Education through Computational Science

An ongoing US National Science Foundation-funded project, the Institute for Chemistry Literacy through Computational Science (ICLCS), involves instructing rural Illinois high school chemistry teachers in the educational uses of computation. The project's training helps participants become educational leaders and change agents.

Education, Outreach, and Training for High-Performance Computing

Most current models for delivering HPC education, outreach, and training lack support for the next generation of HPC users. The authors present lessons learned in developing and disseminating HPC-enabled tools and curriculum, including their efforts at developing low-cost hardware, software, and curricular platforms for easy classroom and workshop deployment.

Community-Cyberinfrastructure-Enabled Discovery in Science and Engineering

A community cyberinfrastructure would enable a new era of multidisciplinary research and collaboration in science and engineering. With such an infrastructure, researchers could share knowledge and results along with computing cycles, storage, and bandwidth. A generic, transparent cyberinfrastructure would also foster more meaningful analyses of data and visualization, modeling, and simulation of real-world phenomena.

Solving Complex Computational Problems Using Multiagents Implemented in Hardware

Using multiagents in reconfigurable hardware can solve complex problems quickly. This new paradigm has the potential to greatly increase computational systems' flexibility, efficiency, and expandability and provides an attractive alternative to current agent-based approaches.

High-Dimensionality Data Reduction with Java

The author used the Java Multimedia Framework along with statistical software from the Colt project to preprocess and compare videos downloaded from Internet sites. Preprocessing the videos decreases each one's dimensionality and size, making it easier to analyze, interpret, and convey useful information about the data's most relevant attributes.

GPULib: GPU Computing in High-Level Languages

GPULib helps scientists and engineers take advantage of GPUs from within high-level programming environments without requiring any detailed knowledge of the GPU architecture.

Why Computer Architecture Matters: Thinking through Trade-offs in Your Code

This three-part series shows how applying knowledge about the underlying computer hardware to the code for a simple but computationally intensive algorithm can significantly improve performance. The final installment focuses on modifying a specific algorithm by applying general principles of efficient programming.

This Time for Sure

The emergence of multicore, heterogeneous architectures for computer games has put parallel computing back in the limelight. The problem is, almost nobody knows how to program these parallel computers for top performance. Research groups developing parallelizing languages and compliers are saying, "This time for sure," but will these tools really help in any dramatic way? Francis Sullivan argues that the central issue is the algorithms.