ITHACA, N.Y. Cornell Theory Center has put together a multi-supercomputer system aimed at discovering how defects form and propagate through solids.
The center's Digital Material Environment uses Python, a high-level scripting language, to access supercomputers and software packages over a network to implement a simulation.
One line of attack on the problem of defect formation will start at the atomic level, studying regular lattice arrangements and their behavior under stress. Another tack will start from observed macroscopic defects and work down to the atomic level.
The work might help device and circuit designers better understand the underlying material dynamics that limit such exotic devices as multiple-quantum-well transistors, which degrade as a result of lattice strain. The new knowledge could also be applied to scaling interconnect on dense VLSI circuits.
The National Science Foundation's Knowledge and Distributed Intelligence Initiative chipped in $1.5 million to get the project off the ground.