High Performance Computing for Nucleic Acid Nanotechnology
Abstract
Abstract Major research thrusts of our group that are currently funded by the ONR and ARO include the development and worldwide deployment of an online server for the automated design of nucleic acid-based nanostructures (CanDo: http://cando-dnaorigami. org) and the application of this framework to understand basic principles of nanoscale multi-enzyme reaction cascades and energy harvesting for the optimization (http://muri.biodesign.asu.edu). The acquisition of a high performance computing cluster in the previous funding cycle has greatly impacted research in the group, allowing very large assemblies to be analyzed with our current framework and gain detailed chemical insights from fully atomistic simulations of nucleic acid structures in explicit solvent and ion conditions. We seek here to expand our current system by adding additional computing nodes and augmenting all nodes (new and existing) with massively parallel NVIDIA GPU processor cards. The expanded HPC server will continue to serve the educational mission of MIT and the DoD by their use in graduate courses taught by the PI, in which students carry out in-depth computational research projects utilizing the servers, as well as by students enrolled in the Computational & Systems Biology (http://csbi.mit.edu/education/phd.html) and Computation for Design & Optimization (http://computationalengineering.mit.edu/education/) graduate programs. The cluster additions we seek to purchase are 9 more computing nodes and 34 NVIDIA Kepler cards. This will raise our effective compute core count to 660 simultaneous process streams with high memory access to perform complex optimization, finite element, and molecular dynamics analyses. The nodes will be accommodated in our existing InfiniBand network that is necessary for high-memory finite element calculations and maintenance of large optimization data. This connectivity allows for multiple compute nodes to be used simultaneously for large jobs, as well as dedicated portions of the cluster to be devoted to our Worldwide DNA Nanotechnology server CanDo and our educational mission at MIT.
Document Details
- Document Type
- DoD Grant Award
- Publication Date
- May 22, 2016
- Source ID
- N000141512830
Entities
People
- Mark Bathe
Organizations
- Massachusetts Institute of Technology
- Office of Naval Research
- United States Navy