Expertise

Creating value for our customers through state-of-the-art expertise.

Data Analysis &
Scientific Visualization

'Applying a Lean approach to Data Analysis & Visualization, embracing uncertainties and complexities for maximum value creation to our customers'

Data Analysis Infrastructure Audit
Software Architecture Review
Proposition and Development of Strategic Technology Direction adapted to our client's needs and requirements

Patents

P.P. Pébaÿ, J.M. Brandt, A.C. Gentile, D.J. Hale, D.C. Thompson, and Y.Marzouk. System and method for statistically monitoring and analyzing sensed conditions. United States Patent No. 7,756,682 B1, July 2010.More...

Selected article

J. Levine, D. Thompson, J.C. Bennett, P.-T. Bremer, A. Gyulassy, V. Pascucci, and P.P. Pébaÿ. Analysis of uncertain scalar data with hixels. Scientific Visualization, 35-44. Springer, 2014, ISBN 978-1-4471-6497-5. More...

Parallel Algorithms R&D

Selected articles

P.P. Pébaÿ. A Novel Shared-Based Approach for Asynchronous Many-Task Models for In Situ Analysis. SC17, ISAV 2017: In Situ Infrastructures for Enabling Extreme-Scale Analysis and Visualization workshop, Denver, CO, U.S.A., September 2017 More...
P.P. Pébaÿ, T.B. Terriberry, H. Kolla, and J.C. Bennett. Numerically stable, scalable formulas for parallel and online computation of higher-order multivariate central moments with arbitrary weights. Computational Statistics, 31(4):1305-1325, 2016. More...
P.P. Pébaÿ, J.C. Bennett, D. Hollmann, S. Treichler, P. McCormick, C. Sweeney, H. Kolla, and A. Aiken. Towards asynchronous many-task in situ data analysis using Legion. Proc. 30th IPDPS, High Performance Data Analysis and Visualization Workshop. IEEE, Chicago, IL, U.S.A., May 2016 More...

Patents

J. Mayo, D.C. Thompson, and P.P. Pébaÿ. System and method for polytopic mesh refinement. United States Patent No. 8,274,512 B2, September 2012.More...

Reduced Order Modeling

Selected articles

Physics-based simulation are playing an increasingly important role in science and engineering. As computing platforms become steadily more powerful, more emphasis is being put on model fidelity and uncertainty quantification (UQ)-based predictive science.
High fidelity simulations are typically characterized by fine spatio-temporal resolutions, leading to large-scale models whose simulation time can take several days or months on thousands of computing cores. Furthermore, the use of UQ techniques for failure, risk assessment and design is taking a significant role within the domain of predictive modeling.
These problems typically require the target model to be simulated thousands of times under different operating conditions, which can become impractical for large-scale simulations. Reduced-order models (ROMs) are a promising technology that can help break this computational barrier.
These methods use simulation data and subspace projection to reduce the dimensionality of computational models while preserving fundamental physical properties such as global conservation and stability.

Computational Geometry

Selected articles

G. Harel, J.-B. Lekien, and P.P. Pébaÿ. Lean Visualization of Large Scale Tree-Based AMR Meshes. SC17, The 2nd International Workshop on Data Reduction for Big Scientific Data (DRBSD-2), Denver, CO, U.S.A., September 2017. More...
G. Harel, J.-B. Lekien, and P.P. Pébaÿ. Two new contributions to the visualization of AMR grids: I. interactive rendering of extreme-scale 2-dimensional grids II. novel selection filters in arbitrary dimensions. CoRR, abs/1703.00212, March 2017. More...

Software Expertise

Languages: C, C++ (98/03/11/14), Python (2/3), QML, MATLAB/Octave, LaTeX

Programming systems and libraries: VTK, MPI, DARMA, OpenMP, PBS

Software process: CMake/CTest, Jenkins, GitLab CI

back