Klaus Schulten, James C. Phillips, Laxmikant V. Kalé, and Abhinav Bhatele.
Biomolecular modeling in the era of petascale computing.
In David Bader, editor, Petascale Computing: Algorithms and
Applications, pp. 165-181. Chapman and Hall/CRC Press, Taylor and Francis
Group, New York, 2008.
SCHU2008
Each time step in a biomolecular simulation is small, yet we need many million of them to simulate a small interval of time in the life of a biomolecule. Therefore, one has to aggressively parallelize a small computation with high parallel efficiency. The NAMD design is based on the concept of Charm++ migratable objects and is fundamentally adequate to scale to petascale machines--this is indicated by the 1-2 milliseconds time per step achieved by NAMD for some benchmarks, with ratio of atoms to processor in a similar range that we expect to see on petascale machines. We have demonstrated scalability to machines with tens of thousands of processors on biomolecular simulations of scientific importance. Implementation strategies have been reworked to eliminate obstacles to petascale through memory footprint reduction and fine grained decomposition of the PME computation. All this has made the study of large molecules such as the ribosome and entire viruses possible today and will enable even larger and longer simulations on future machines.
Download Full Text
The manuscripts available on our site are provided for your personal
use only and may not be retransmitted or redistributed without written
permissions from the paper's publisher and author. You may not upload any
of this site's material to any public server, on-line service, network, or
bulletin board without prior written permission from the publisher and
author. You may not make copies for any commercial purpose. Reproduction
or storage of materials retrieved from this web site is subject to the
U.S. Copyright Act of 1976, Title 17 U.S.C.
