Page not found

Modeling Transition Metal and Actinide Chemistry

The research group of Professor Laura Gagliardi ( Chemistry ) uses quantum chemistry methods to study chemical systems containing transition metals and even heavier atoms like lanthanides and actinides. Gagliardi group projects that currently use MSI include: simulations of actinide chemistry in...

Enthought Canopy Python Distribution

Python is a high level programming language that aims to combine remarkable power with very clear syntax. The Enthought Python Distribution is a cross-platform environment for scientific computing in Python, and includes the Canopy IDE and package manager. MSI has installed an academic-licensed version that includes hundreds of modules, including tools that enable efficient parallel computations.

MSI Upgrades Storage Solutions

One petabyte of tape storage for under $10,000? Leveraging new advances in tape media, the Minnesota Supercomputing Institute has expanded its storage portfolio to offer easy access to deep archival storage at low cost. MSI currently offers a high-performance storage solution from Panasas (3.2 PB...

Statistical Models for Dependent High-Dimensional Data

Abstract: 

Statistical Models for Dependent High-Dimensional Data

These researchers are involved in several projects using MSI.

  • Statistical Methods for Spatial Data: This research on spatial models has focused on regression inference for areal, i.e., spatially aggregated, data. Areal data are common in many fields, including forestry, marketing, epidemiology, image analysis, and ecology. Since investigators in these fields are often interested in scientific explanations rather than, or in addition to, predictions, spatial regression is important.
  • Spatiotemporal Inference for fMRI Data: Typical fMRI experiments generate large datasets that exhibit complex spatial and temporal dependence. Fitting a full statistical model to such data can be so computationally burdensome that many practitioners resort to fitting oversimplified models, which can lead to lower quality inference. These researchers have developed a full statistical model that permits efficient computation.
  • Joint Models for Longitudinal Data: The researchers have developed semiparametric and nonparametric joint models for multidimensional longitudinal outcomes. Although they focus on revealing time-varying dependence relationships, the frameworks accommodate all manner of time- varying parameters for the coordinate processes: regression coefficients, variances, etc. These methods will allow researchers to reveal complex dynamic patterns of dependence and response–predictor relationships.
  • Piecewise Growth Mixture Models: This project focused on piecewise growth mixture models (PGMM), a special case of the finite mixture of multinormals. The researchers investigated Bayesian inference for PGMMs and maximum likelihood inference by expectation maximization.
  • Bayesian Inference for Gaussian Copula Regression Models: Gaussian copula regression models (GCRM) provide a flexible, intuitive framework for modeling high- dimensional dependent outcomes. When such outcomes are discrete, the likelihood is computationally intractable because the running time grows exponentially in the sample size. These researchers developed three computationally feasible approaches to Bayesian inference for GCRMs with discrete outcomes.

Return to this PI's main page.

Group name: 
hughesj

High Throughput Screening Data Analysis

Abstract: 

High Throughput Screening Data Analysis

The High Throughput Screening Facility supports tool and drug discovery in the Institute for Therapeutics Discovery and Development in the College of Pharmacy at the University of Minnesota. The High Throughput Screening Team conducts biochemical and cell-based assay development, high throughput screening, and structure-activity relationship studies across multiple therapeutic areas. This researcher uses ActivityBase, SARview, Pipeline Pilot, and Spotfire to analyze the high throughput screening data generated in his laboratory. This involves analysis of datasets containing up to 250,000 compounds, and involves Hit Identification, analysis of the performance of the screening campaign, preliminary structure-activity relationships, and analyses of the physico-chemical properties and structural similarity of the Hits and Hit families. Additionally, analysis will be conducted of the screening collection of ~250,000 compounds including, for example, diversity and cluster analysis, as well as analysis of the impact of additional new screening collections on these properties of the full compound collection.

Return to this PI's main page.

Group name: 
hawkinso

First-Principles Modeling of Transition-Metal Minerals and High Tc Superconductors

Abstract: 

First-Principles Modeling of Transition-Metal Minerals and High Tc Superconductors

During 2015 this group will use MSI resources to perform first-principles calculations (based on DFT) on two main classes of materials i) Fe-bearing minerals (e.g., Fe-Mg-Al perovskites and postperovskites) at the pressure and temperature conditions typical of the Earth interior; and ii) high-Tc layered cuprates (e.g., La2CuO4). The first project will investigate the thermo-elastic properties of Fe-containing minerals at the conditions of the earth's interior. The second will focus on the magnetic and vibrational properties of the parent (undoped) materials and on the role of electron-phonon interactions in determining the finite-temperature properties (e.g., the superconductivity) of doped systems. Calculations will be based on state-of-the-art DFT functionals with newly-developed corrections to improve the treatment of electronic correlations. 400,000 SUs are requested in total to develop these research projects. 200,000 SUs are requested for calculations on minerals, 200,000 SUs will be used for the study of high-Tc cuprates.

A bibliography of this group’s publications is attached.

Return to this PI’s main page.

 

Group name: 
cococcio
Attachment: 

Novel Paradigms in Geometric Modeling of Large and High-Dimensional Datasets

Abstract: 

Novel Paradigms in Geometric Modeling of Large and High-Dimensional Datasets

These researchers aim to develop effective data modeling paradigms that are sufficiently simple for statistical inference. Current scientific investigations, as well as industrial applications, produce and rely on massive, high-dimensional and possibly corrupted datasets. A major focus of applied mathematicians and statisticians in this area has been on quantitative geometric data modeling. In order to effectively analyze large data and obtain meaningful statistical inference, the underlying geometric models need to be sufficiently simple. This project suggests mathematical paradigms for such effective geometric models. It plans to develop rigorous mathematical theory for these paradigms combined with carefully designed numerical strategies addressing specific and important applications. Despite the recent progress in this area, there are many open directions, several of which this project addresses.

More specifically, this research focuses on several important directions of geometric data modeling. One direction aims to address modern issues in single robust subspace modeling with respect to new paradigms of learning and computation that have hardly been addressed so far in this setting. Another direction will explore important issues in modeling data by multiple subspaces or manifolds with new paradigms and perspectives. The project will also emphasize specific paradigms of low-rank and sparse modeling, which are induced by important applications, such as approximate nearest subspace for object recognition, improved feature tracking, structure from motion in computer vision, and sparse modeling in the atmospheric sciences.

Return to this PI's main page.

Group name: 
lermang

MSI 2014 Research Exhibition

posted on April 28, 2014 MSI held its Fifth Annual Research Exhibition on April 24, 2014. This event showcases research being performed using MSI resources and includes a poster competition. We had 44 posters submitted to the Exhibition this year, in a wide variety of disciplines. Posters were...

Multiscale Design of Hard and High Temperature Resistant Coatings

Abstract: 

Multiscale Design of Hard and High Temperature Resistant Coatings

This project involves an interdisciplinary effort to conduct multiscale design of hard and high temperature resistant (Si,Zr)-B-C-N coatings which are thermally stable and oxidation resistant for high temperature (>1500 °C) applications. The project couples multiscale computations and experiment to merge the high-temperature oxidation resistant properties of Si-B-C-N and high hardness properties of Zr-B-C-N systems. The predictive effort spans from atomistic to multiscale distinct element method simulations to formulate solid predictions of the optimized compositions. These predictions will provide critical guidance for synthesizing coatings with targeted properties. These researchers expect that in these new coatings, the desirable properties will coexist, resulting in a new generation of protective layers.

This research is far-reaching as it can enable new concepts for protective coatings and the development of a new multiscale tool to predict materials' response. Molecular dynamics investigation will address the fundamental issue of combining desirable properties by varying chemical composition and structure. The application of the distinct element modeling down to the nanoscale represents a new powerful tool to simulate the global behavior, allowing the design of future materials at large. The focus of this research - the discovery of new coatings working under extreme conditions - can find application in multitude of critical components such as turbine blades, reusable launch vehicles, hypersonic vehicles, and thermal barrier applications.

Research Spotlights about this group's work appeared on the MSI website in August 2014 and March 2015.

Return to this PI’s main page.

Group name: 
dumitric

Simulation of High-Speed Turbulent Combustion

Abstract: 

Simulation of High-Speed Turbulent Combustion

Hypersonic air-breathing propulsion systems were first envisioned about six decades ago. The successful flights of the NASA X-43 and the recent partial success of the Air Force X-51A demonstrate that these systems can work. The engines in these vehicles are mechanically extremely simple, compared to turbofans or even automobile internal combustion engines. However, the coupled fluid dynamics and chemical energy conversion that takes place inside these engines is anything but simple. Almost every non-linear fluid dynamics phenomenon plays a role in their operation: turbulent boundary layers, free shear layers, shock waves, fuel-air mixing, and finite-rate chemical reactions all interact with one another in an extremely dynamic and energetic environment. The Candler research group is developing novel large-eddy simulation (LES) approaches to simulate these complex flows. LES simulations resolve the large-scale unsteady turbulent motion and model the unresolved subgrid-scale motion. This approach has been shown to correctly represent complex geometry turbulent motion for a wide range of applications. However, appropriate numerical methods and subgrid-scale models have not been developed and validated for the highly compressible conditions that characterize the high-speed combustion systems. The group has recently developed a novel subgrid modeling approach, along with associated improvements to the numerical methods. Thus, they are using MSI computer resources to simulate experimental configurations and compare the simulations with the available experimental data.

Return to this PI's main page.

Group name: 
candlerg

Pages