Twenty-five years as a professor and scientist at the forefront of computational fluid dynamics research have given ICME's Founding Director Parviz Moin invaluable insights into how computational engineering has evolved and where it's headed. He shares some of them below.

Research Tools of the Future

"My focus really is on application" continues Sheppard. "I want students to know how to open the"Originally, scientists' research methodologies were theory and experiment," he says. "Computing now forms the third leg of a stool for scientific discovery and prediction, and for studying the performance and behavior of physical systems."

Until very recently the complexity of multi-scale problems exceeded the computational resources of all but a handful of labs. Computers were too slow to handle the data effectively enough to apply to real-time problems. "Today, however, we have extremely fast computers and parallel processors to do large-scale simulations of aspects of the real world, from jet engines to the cardiovascular system in the human body," says Moin. "Computers make it possible to visualize the results and do it quickly. These simulations create virtual laboratories. In fact, computers are our laboratories and our primary tools, and they are increasingly accessible to engineers working in industry."

New Multi-Disciplinary Program Suited to Professionals

"The modern computational engineer must have a broad understanding of different areas of computational math: multi-scale modeling, multi-physics, discrete mathematics and optimization," Moin says. "New computational tools require additional training to effectively use them; this is crucial. Today's engineer must be competent with efficient use of supercomputers. One may know how to run a program for an application, but if he or she cannot use it efficiently, this is practically useless."

The CME curriculum builds on Stanford's Scientific Computing and Computational Mathematics program. "It has a broader scope," explains Moin. "The master's program integrates continuous and discrete math with computation. There are also new programming classes for modern computational engineering Our curriculum committee brings professors who actually teach the subjects together regularly to discuss how to do this best. We have designed the MS degree program to be very well-suited to industry professionals-that is, one may succeed and benefit from the program without being dependent on the on-campus culture of ICME."

New courses developed for ICME include service courses for first-year master's students, core courses for master's and doctoral students and specialized electives. "These have been designed to give master's and doctoral students a toolbox of mathematical and computational methods for solving problems in their specific fields."

ICME faculty have looked beyond the master's degree program, recognizing the increasing need for a solid CME background in many engineering fields. "Service courses offer a balance of both discrete and continuous math courses that fulfill the math requirement for a variety of engineering master's degree programs," Moin explains.

Finally, Moin points out that Stanford has distinct assets to offer that are reflected in ICME. The university has world-class programs in many areas of engineering and other disciplines and actively promotes interdisciplinary cooperation and research. "Graduate students have the advantage of being able to work with leading faculty throughout the university," he says.

The Right Fit: Ability, Passion for Computers and Problem-Solving

Moin teaches CME206: Introduction to Numerical Methods for Engineering (Was ME300C) every spring, and he understands the challenges for working professionals who pursue degrees at a distance. "Because our tools are computers, the curriculum is appropriate for distance students at the master's level. A problem is, however, that students who are working are sometimes pulled away by the demands at their job sites," he says. "If they are taking courses online, it also can be very difficult to work in isolation. I encourage them to take courses as enclaves whenever possible."

What sort of student will get the most out of the CME curriculum? "We have strong emphasis on requiring our students to be good programmers," he says. "Remote student must not only be interested in and have ability in mathematics; they must also be excited about using computers to solve problems."

Moin envisions Stanford's ICME as an industry resource for the future. Industry students have the potential to be a conduit to productive research and industry partnerships. "I think that the students might introduce modern computational mathematics into their companies," he says, "which could lead to identification of real-world problems that a multi-disciplinary institute like Stanford could help with."

### For more details on the CME MS degree program, faculty spotlights and more, visit and bookmark http://scpd.stanford.edu/SCPD/courses/contentView/icme/.