Andrew Christlieb: A large-scale step into the world of computational science

  • Apr 14, 2015
  • Homepage Images, News

Andrew Christlieb is an MSU Foundation Professor of mathematics and chair of MSU’s new Department of Computational Mathematics, Science and Engineering, which is jointly administered by the College of Natural Science and the College of Engineering. 

The new Department of Computational Mathematics, Science and Engineering (CMSE) is the result of a long-term effort to raise the level of research, graduate and undergraduate education at MSU.  A little more than four years ago, a group of NatSci and College of Engineering professors proposed a graduate program in computational mathematics, with the goal of filling MSU’s void in large-scale scientific computing. When it immediately became evident that there would not be enough faculty on campus to support such a program, the next step was to consider an institute. But it soon became clear to us that the science of developing fundamental algorithms for solving problems in science and engineering has turned into a field in its own right. So we started advocating for a place where faculty could have a tenure home—a department dedicated to computational science.

Two years ago, we began the major thrust in putting the proposal together for the new department. The goal in NatSci was to break down silos, and provide interconnectivity between disciplines. Maybe not break the silos down—maybe drilling holes into them is the right analogy!

MSU has tremendous expertise in a wide range of research areas, including nuclear physics, plant biology and energy-related materials. But what’s been missing is the computational aspect.

Large data is coming up everywhere within science. The problem is, the data sets are growing in size faster than our computing technology is growing. The question becomes: “How are we going to develop tools for analyzing this kind of data that lets us better understand the world around us?” We can’t hope to understand data unless we have models to help us. MSU’s new department will provide the necessary computational arm that will design new algorithms and develop tools for investigating new models. We want this department to be about the algorithm of scientific computing, but to have bridges to many disciplines that are critical on campus.

The department will consist of two distinct areas—foundations of data science and scientific computing. Further, the department will be a blended environment where we provide a home for a range of scientists who focus on a range of problems, spanning everything from theory to application. It is this second aspect of the new department that will make it a dynamic setting for doing research. Data science focuses on how you make sense of crazy amounts of data (big data is always relative; what’s big data today will not be big data next year because computers will change). In the area of scientific computing, or large-scale computing, what we’re interested in is developing algorithms that are going to allow us to solve large problems in an efficient, cost-effective way in this crazy new world of multi-core computing; and it’s only going to become more entrenched as we go forward. A serious amount of money is invested in running these computers. Just turning on a supercomputing system could cost around $10,000; running the calculations for several days could cost tens of thousands of dollars.

In the past 10 years, 37 departments focusing on either scientific computing or data science have popped up at universities nationwide; it’s just exploded. Within the schools that are ranked in the top 100 by the National Research Council, 9 have departments that focus on scientific computing and 6 have departments that focus on developing methods for challenging problems in data science. Only one of these universities—Georgia Tech—has created a school that has substantial expertise and degree programs in both of these fields.

In order for MSU to be at the forefront of scientific discovery through computation, we knew that now was the time to take a worldwide leadership role by forming a new interdisciplinary department focusing on the intersection of algorithm, application, and data.

Currently, 17 faculty members from NatSci and the College of Engineering are involved at various levels of participation in developing the new department, with 6 who are moving part of their appointments into the new department. This year we hired our first 9 new faculty related to the new department, with 5 of them having a majority appointment and 4 of them having a minority appointment in CMSE. We’re hoping to hire 8 additional faculty next year, 8 the following year, and 6 the year after. This would bring us to what I would call teaching capacity, or educational capacity. We plan to launch our PhD program in fall 2016 and our undergraduate degree programs by 2018. At full strength, we anticipate having 3-5 graduate students per FTE. We will also be launching a minor, which would support 200-300 students.

We are getting extremely positive vibes from industry about what we’re doing. When we talk to companies like Procter & Gamble and Dow about what we’re proposing for the majors, their immediate response is: “When can the students start?” The problem is, we can’t generate the students fast enough. Representatives from industry remain very enthusiastic; and we are excited about that!