Biological molecules and their associated phenomena are crucial to life as we know it. The ability to simulate such phenomena would allow for deeper fundamental understanding of basic life processes, help elucidate diseases such as Alzheimer's, and offer the promise of reduced costs and time in fields such as drug discovery. Paradoxically given their small size, detailed physical simulation of biomolecules requires massive amounts of computational power. This, along with other factors, has long limited the problems that could be addressed.
Distributed computing and new computational technologies can help address the challenge. Now with over 200,000 computers participating worldwide, the [email protected] project is more powerful for parallel simulation than any supercomputer in the world. With the recent addition of Sony PS3 participation, it has a sustained performance of over 800 TFLOPS. This talk will describe advances in the technological infrastructure of the project, and describe examples of methods that have been developed to take advantage of parallelism.
Guha Jayachandran is completing his Ph.D. in computer science (Chemistry minor) at Stanford University under Prof. Vijay Pande. He will graduate in June. Guha's interests include computational drug discovery, novel applications of biocomputation, and grid computing. He has worked on all aspects of [email protected] -- a worldwide distributed computing project examining protein biophysics -- for over five years, also focusing during recent years on the development of new computational methods for probing biomolecular processes like drug binding.
Official Website: http://sfbayacm.org
Added by marstein on June 18, 2007