About Us

Background

Research Focus

People

Repository

Workshops

Projects

Links

Contact MRI

CCS: Institutes: MRI: Background: Enabling National Science Goals

• Introduction
• Scientific Rationale
• HPC Rationale
• Thrust of CCS-MRI
• Resources of CCS
• Next Steps
• Enabling National Science Goals

Enabling National Science Goals

During the last two decades advances in computational materials and condensed matter research has undergone a revolution in the complexity of physical phenomena that have become amenable to theoretical modeling and simulation. In materials science, theory, modeling, and simulation are increasing in a symbiotic relationship with experiment in the scientific discovery process. Here CCS-MRI is poised to play a decisive role in enabling major national initiatives.

CCS-MRI implements the central goals of the SciDAC initiative within computational materials science nationally by supporting:

     "Scientific Challenge Codes - research, development and deployment of mathematical models, computational methods and scientific research codes to take full advantage of the capabilities of terascale computers … and (sic) solve critical scientific problems ..."

     "Computing Systems and Mathematical Software - development and deployment of software to accelerate the development of and protect long-term investments in scientific codes, to achieve efficiency on high end computers … "

     "Collaboratory Software Infrastructure - … deployment of software to link geographically separated researchers …."

Though these devices CCS-MRI can pay a key role in enabling the materials research scientific goals of Office of Basic Energy Sciences (BES) Division of Materials Sciences and Engineering (DMSE) programs. Such as the National Nanoscience Initiative (NNI) as well as in support of the computational infrastructure and resources required to take full advantage of major BES-DMSE funded user facilities such as the the Spallation Neutron Source (SNS).

The CCS-MRI has a potentially unique role to play in nanoscience. For example, the software environment and computational resources available within CCS-MRI will allow direct simulation of nano-structures that are the subject of experimental investigation - this is a new world. Nanoscience is where theory, simulation, experiment, and realizable technology most starkly confront each other and where modeling and simulation can have a revolutionary impact on the way materials scientists make discoveries.

Major DOE user facilities such has SNS, IPNS, ALS, and APS probe the properties of matter at ever more detailed levels. The volume and complexity of the data demand sophisticated analysis tools and models in order to extract the full value from these facilties. Again the computational tools and terascale computing facilities of CCS-MRI can play a key role, particularly with respect to SNS, by providing and supporting the software infra-structure and high-end computational resources necessary for scientists using SNS to analyze their data. Interactive high level data tools and modeling could allow rapid analysis and optimization of experimental procedures within a single access cycle.

The various models for Si-rich SiC(001)(3x2) reconstruction are extremely close in energy, but comparison between the measured and the calculated surface optical signal (reflectance difference spectra) allows for identification of the two-adlayer asymmetric dimer model (TAADM) as the dominant structure. [W. Lu, W. G. Schmidt, E. L. Briggs, and J. Bernholc, Phys. Rev. Lett. 85, 4381 (2000).]

The Computational Materials Science Network (CMSN) is a recently formed national virtual network of University, Government Laboratory and Industry researchers funded through BES-DMSE who's mission is to "advance frontiers in computational materials science by assembling diverse sets of researchers committed to working together to solve relevant materials problems that require cooperation across organizational and disciplinary boundaries".

CCS-MRI can play an important role in enabling the scientific goals of CMSN by providing the software infrastructure and dedicated terascale computing environment required to address the frontier research focus areas of CMSN projects.

CCS-MRI presents a new way of approaching computational materials science. The 'cottage industry' approach of the past is replaced by a new paradigm in which a sophisticated materials science tool set is integrated with terascale computational facilities. This will then allow materials scientists to move away from the time consuming, repetitive, and duplicative tasks of developing and maintaining codes to a mode where a software environment is rapidly assembled to solve the problem at hand.

The computational resources made available though CCS-MRI will complement those of the National Energy Research Scientific Computing Center (NERSC) at Lawrence Berkeley Laboratory (LBL)., (see: hhtp://www.nersc.gov). While the HPC capabilities of NERSC and CCS-MRI are similar, NERSC currently supplies the whole DOE ER community. As a result NERSC has a very large user base with only few selected projects having large allocations of HPC resources. In CCS the total resources will be devoted to large-scale projects in the major focus areas of CCS (Materials, Climate, Biology). Consequently the user base will be small, providing individual projects with large total resources and the ability to routinely run jobs of a size not generally available at NERSC. In addition, the software developments undertaken within CCS-MRI will be undertaken in close collaboration with NERSC staff such that tool developments at one site will be automatically available at both sites.