Research

Research Opportunities in Computational Science and Engineering

A full description of the opportunities for developing CSE research at Berkeley would require a volume unto itself, as nearly all fields of science and engineering are now impacted by large-scale computational research. We include a few examples of the computational problems our doctoral students will be able to address.

• Combustion Modeling. Combustion provides more than 85% of U.S. energy, and meeting these needs with lower environmental impact requires more efficient combustion systems, which are being studied through computation. These turbulent combustion computations typically span a huge range of time and length scales (as much as a factor of 109), requiring large and adaptively managed sets of gridpoints.

• Climate Modeling. Understanding climate change is one of the most important problems facing us all. It is a major challenge to include not just more finely resolved versions of the coarse air, sea and land models now in use, but much more detailed models of other biological and human activities that play a large role in the climate. Much more finely resolved models including large amounts of measured data are needed to reduce uncertainties. Modeling is also critical to evaluating (potential) human activities or economic policies intended to mitigate climate change.

• Astrophysics and Astronomy are being revolutionized by the ability to collect vast amounts of observational data that needs to be analyzed, in addition to the possibility of large-scale first-principles simulations. CSE will have an impact in research areas such as large-scale structure and cosmology, galaxy formation and interactions, star formation, stellar evolution and death, and numerical relativity. Large impacts are possible in related fields as well, such as by biologists and chemists examining the origins of life by considering the organic chemistry of the interstellar medium, and by geoscientists studying the origins of the planets by studying the numerous planetary systems recently found orbiting nearby stars

• Engineering Better Search Engines. Better searching of web pages and other large databases will continue to have tremendous socio-economic impact by improving the way we communicate and explore the world. There are many technological challenges in organizing the world’s knowledge and delivering services that make it universally accessible and useful, and to understand it and discover important relationships and meanings, research which search companies are interested in supporting.

• Cognitive Computing and Engineering. Better understanding of the mind will have a huge impact on national interests in science, medicine, and human health and well-being as well as targeting cures for many known brain disorders such as Alzheimer’s disease, Parkinson’s disease, and depression. One Berkeley focus will be in Modeling of the Mind, which includes computer science, engineering and mathematics. Another is neural computation, how neurons process and represent information.

• Parallel Computing. ParLab is a large effort among EECS, Mathematics, Mechanical Engineering, and Music faculty to address multicore computing, i.e., the imminent change of the entire computing industry to embrace parallelism. This is made necessary by physical limits on the ability to continue doubling the clock speed of computers every 18 months, and means that challenges of high performance parallel computing once limited to relatively few scientific and engineering problems are now the problems of the entire computing industry. This means that the computer industry will be investing large sums of money to address problems that directly affect CSE.