A recent history: computation at SEAS

February 10, 2014
A recent history: computation at SEAS

CSE SM student Anita Mehrotra explores the history of IACS, the start of the CSE SM program, and the future for computational science at Harvard 

Harvard is heralded as the oldest university in the United States, but only in 1847 – over two hundred years after its formation – did a school specifically focused on engineering arrive at the campus. Named after founding donor and industrialist Abbott Lawrence, the Lawrence Scientific School was Harvard’s first “major effort to provide a formal, advanced education in science and engineering.” Over time the School grew, merged with Harvard College, became a Division of Applied Sciences and then, in 2007, transitioned into the School of Engineering and Applied Sciences (SEAS). Today, SEAS is home to over 70 full-time faculty, 400 graduate students and 550 undergraduates.

Some argue that as a liberal arts school, “Harvard” and “engineering” are paradoxical terms. How can Harvard possibly compete against technical behemoths like nearby MIT and Princeton, and West Coast giants Stanford and Berkeley? Yet a deeper look at SEAS’ projected path, shaped greatly by Dean Cherry Murray’s vision, reveals an approach that renders this competition moot. SEAS focuses on “harnessing the power of engineering and applied sciences [in order to] address the greatest challenges facing society.” The goal, in other words, is to use cutting-edge technological tools to solve social problems.

SEAS’ focus has largely revolved around developing “T-shaped” individuals: students who have deep knowledge in a discipline, but are also able to collaborate across the boundaries of disciplines, essential to addressing 21st century challenges, and a role that complement the more domain specific advancements done in traditional engineering subfields at other universities.

The Beginning of IACS

It was with these particular goals in mind that the Institute for Applied Computational Science (IACS) was founded in 2010. The mission? To grow an in-house group of faculty and students with skills in applied computational science.

Computational science is often described as a conglomeration of computer science, applied mathematics and statistics, but as with any new field, its definition is constantly being refined. This is especially true with the swift-moving, precipitous growth in intensive-computation technologies. “[G]raduate students will need to know how to leverage computer power to solve human problems and to explore all aspects of nature and life, at scales and speeds that will require novel computational thinking,” noted Dean Murray in 2010, speaking to the high demand for data scientists and statisticians which continues to this day.

Additionally, the idea of a computation-focused institute came about because applied mathematical modeling was being used in many humanities and pure-science graduate labs within Harvard itself. In Chemical Physics and Astronomy, “high performance computing (HPC)” -- including, for example, parallel processing on GPUs -- are regularly used for simulations. The Institute for Quantitative Social Science uses machine-learning techniques to aid national security efforts, a top issue for governments around the world.

In the spring of 2011, a series of core Computer Science and Applied Math courses in IACS began. The new classes include stochastic modeling and optimization, and have grown to include Parallel Computing for Computational Science, taught by Cris Cecka, and Numerical Methods for Advanced Scientific Computing, taught by David Knizevic, both IACS lecturers. These classes, and other IACS-developed courses, feature students from a variety of disciplines.

Under the directorship of Efthimios Kaxiras, John H. Van Vleck Professor of Physics, and Rosalind Reid, current Executive Director at the Council for the Advancement of Science Writing and an IACS Fellow, IACS quickly began to grow from an exciting idea into a critical cog in the SEAS wheel.

Computational Science and Engineering as a Degree

In 2012, a stand-alone master’s degree program in Computational Science and Engineering (CSE) was launched. Ms. Reid concisely summarized the goals of the CSE program as building “deployable knowledge.” Together with a committee that included Dean Murray, Kaxiras, Hanspeter Pfister, An Wang Professor of Computer Science, and Pavlos Protopapas, Science Director of IACS, Reid asked difficult questions of faculty and industry giants alike in order to develop a graduate program. “The approach was very forward-looking: who are the leaders in the next five to ten years? What skills do they have?” Ms. Reid recalls. The industry advisory committee that helped shape the program included the likes of Google, Intel Labs and Goldman Sachs.

The inaugural CSE class of 2013 includes twenty-four graduate students from a wide range of backgrounds with vastly different interests, but with a shared thirst for applied computation and a passion to make a difference. The variety reflects a hope for a fully-integrated University, a development of T-shaped individuals, and a desire to hone the technical skills of tomorrow’s societal problem-solvers.

In 2012, Dr. Daniel Weinstock became Assistant Director of Graduate Studies in CSE, and in August of 2013, Pfister became Director of IACS. Together, they have used continued industry input, student feedback, and new courses (like the new, three-way cross-listed Data Science course) to fill in the details of the program and shape what it will look like for future students. The program will continue to grow with the launch of the two-year Masters in Engineering program this next year.

What’s Next? A Look at Further Beyond

In 2018, SEAS will move across the Charles River to Allston, next door to Harvard Business School (HBS), giving SEAS more physical space and collaborative opportunities with MBA students and the i-Lab. Ultimately, such a move will further the goal of a more fully integrated university. Because of its unique position at the intersection of computer science and applied mathematics, IACS and the CSE programs will be impacted most of all – and will likely leave a big impact on HBS, positioning computational science to change the university as it has already started to change the outside world.