Computation at Cornell
Through generous donations by Intel Corporation, the Cornell Nanoscale Facility (CNF) has provided high performance computing capabilities to users since early 2005. This effort is part of the greater computational initiative in the NNIN to provide nanoscale modeling resources that accelerate research and innovation. The Intel cluster consists of 152 Xeon processors (288 computing cores) linked with Gigabit connections. A portion of the cluster is also linked by Infiniband fabric for intensive parallel calculations. The cluster hosts an ever-expanding and diverse suite of simulation tools for nanoscale systems including codes for first principles calculations, photonic devices, molecular dynamics, multiphysics, and nanoscale transport.
While numerous standard packages exist for well-known systems in different fields, cutting edge research often requires developing new algorithms or approaches to address unique problems. To this end, the CNF is dedicated to not only providing simulation tools, but also playing a role in their development. Dr. Derek Stewart serves as the scientific computation research liaison for the NNIN at the CNF and works with users to modify existing codes or need requires, constructing new approaches. The cluster also serves as a crucial test-bed where codes developed in research groups can be tested by users and develop the robustness necessary for wide spread distribution.
For more information, please contact Derek Stewart at firstname.lastname@example.org.
- Electronic Structure of Nanoscale Systems (Tight Binding, Density Functional Theory, Monte Carlo Approaches)
- Nanophotonics Simulation Tools
- Molecular Dynamics
- Modeling Nanoscale Electronic and Thermal Transport
- Quantum Chemistry
- Cyberinfrastructure Tools for Research: (Virtual Vault for Pseudopotentials)
Simulation Workshops hosted at the CNF
The CNF has hosted and organized numerous workshops that also provide training sessions on different simulation tools.
Lectures, tutorials, and some videos on many of the codes available can be found at:
- 2005 CNF Fall Workshop "Modeling the Nanoscale World"
- 2006 CNF Fall Workshop "Building Nanostructures Bit by Bit"
- 2007 CNF Fall Workshop "Defining the Interface between Nanoscience and Geology"
- 2010 NNIN/NCN Joint Fall Workshop "Building a Collaborative Framework for Nanoscience"
- 2012 PASI Workshop, Santiago, Chile "Computational Material Science for Energy Generation and Conversion"
Pubications Resulting from the CNF Cluster
Computational Nanotechnology Codes Supported at CNF
Density Functional Approaches
- Abinit - an open source robust plane wave density functional code
- Wien2k - Electronic structure calculations using density functional theory based on a full-potential linearized augmented plane wave approach.
- LM Suite - Linear Muffin Tin Orbital package that support full potential and atomic sphere approximation (ASA) calculations. It can also model electronic transport in nanoscale structures using a non-equilibrium Green's function approach.
- PARSEC - (Pseudopotential Algorithms for Real Space Energy Calculations) solves density functional calculations using a real space approach. This technique is ideal for modeling small clusters.
- CPMD - Car-Parrinello Molecular Dynamics - This code can be used to perform ab-initio molecular dynamics calculations. It allows for time-dependent density functional calculations, wavefunction optimization, and path-integral molecular dynamics.
- Siesta - (Spanish Initiative for Electronic Simulations with Thousands of Atoms). This code using numerically truncated orbitals (single and double zeta approach) to build an order-N density functional code. This code is ideal for modeling large scale nanostructures (i.e. nanotubes, nanowires, molecules).
- Plato - (Package for Linear-combination of ATomic Orbitals) is a suite of programs designed and written by Andrew Horsfield and Steven Kenny. Capabilities of the program include a tight-binding algorithm, and a self-consistent field method using either a local density approximation or generalized gradent approximation functionals. The program was specifically developed for application to systems with periodic boundary conditions (crystals), however it is also possible to treat isolated molecules. PLATO was written with speed of calculation as a priority, and uses numerical basis sets and tables of pre-computed one- and two-centered overlap integrals.
- Quantum Espresso - (also known as PWscf) This plane wave density functional code takes advantage of ultra-soft pseudopotentials to accelerate calculations. In addition, it has the ability to handle magnetic structures, calculate phonon dispersions, and perform structural relaxations.
- Fleur - This code is based on the all-electron full potential linear augmented plane wave approach. It can provide important check for plane wave calculations and also has special options for handling surfaces and 1d structures.
- NWChem - This code is a highly parallel quantum chemistry package capable of Self Consistent Field, Hartree Fock (RHF, UHF), and Gaussian density functional calculations as well as several other techniques.
- QuantumWise (formerly Atomistix) ATK and VNL - This commercial package provides the ability to calculate I-V curves in nanostructures and molecular junctions. The code is based on a non-equilibrium Green's function approach and it also takes advantage of numerically truncated orbitals to allow for calculations of large systems.
- LAMMPS - general purpose molecular dynamics simulator that has the option to use leonard jones potentials, embedded atom potentials, and potentials for biomolecules and proteins. This parallel code can easily handle systems with thousands of atoms. The ability to incorporate the effect of temperature provides an important complement to density functional techniques.
- DL_POLY - DL_POLY is a parallel molecular dynamics simulation package developed at Daresbury Laboratory by W. Smith and T.R. Forester under the auspices of the Engineering and Physical Sciences Research Council (EPSRC) for the EPSRC's Collaborative Computational Project for the Computer Simulation of Condensed Phases (CCP5) and the Molecular Simulation Group (MSG) at Daresbury Laboratory. You can find a local tutorial on DL_POLY here and this code was also discussed at the 2007 CNF Fall Workshop.
- GROMACS - this molecular dynamics code was designed primarily to examine biochemical molecules such as proteins, lipids, and nucleic acids. It can also be used to simulate proteins. The authors have done considerable work on optimizing the code.
- MPB - MIT Photonic Bands package. This code can calculate the band structure and electromagnetic modes of periodic dielectric structures.
- MEEP - This is an open source finite difference time domain (FDTD) simulation code that was developed at MIT.
- UTQUANT - This is a quasi-static CV simulator for one dimensional silicon MOS structures.
- SEMC-2D - Schrodinger Equation Monte Carlo 2D simulator for quantum transport and inelastic scattering effects in nanoscale semiconductor devices such as nanoscale double gate MOSFETs and tunnel injection lasers. A dissertation providing more details on the approach is available here
- Elmer - This multiphysics packages allows you to model coupled problems. This could include current induced heating, vibrations in cantilevers, and fluid flow in microchannels.