|Principal Investigator(s)||J.H. van Lenthe|
|Leading Institution||Utrecht University|
|DEISA Home Site||SARA|
We propose to explore the boundaries of farms of Supercomputers in the Quantumchemistry domain, both using taskfarming and to aid farmaceutical sciences.
The limit of current Supercomputers seems to be reached at biological enzymes of some 15 Å in diameter. To establish the validity of current approximations an upgrade in size is still required. The results will be used to benchmark current approached in computational drug design.
The use of task farming for high throughput computational screening of possible catalysts has already firmly established as a viable technique on single multi-processor systems. The number of molecules screened is, however, still only the bare minimum. Distributing the computational load over a wider network might be the solution if the communication overhead does not become prohibitive. An initial test of this concept is now possible.