Stanford Mechanics and Computation

Home page | CV | Teaching | Research | Publications | Students

Research projects

Last update: February 2012

Prof Darve’s group works on fast numerical methods and high-performance computing. Many engineering problems are computationally very expensive to solve. Those can range for turbine engine simulations to red blood cells and nanometer-sized molecules. In all these cases, even the largest computers are insufficient to simulate these systems because of the computational cost. These challenges can be addressed by developing fast algorithms and making use of state-of-the-art computing hardware such as graphics processing units (GPUs). As an example, we developed an algorithm that can perform an LU factorization of a matrix of size 500,000 in 29 secs whereas a traditional method takes more than 30 days. Current application examples include carbon sequestration, red blood cells and platelets, and proteins.


Barium Titanate Crystal


ASIC1A ion channel protein


GPU for high-performance computing