Lion-codes

GPU research and teaching applications

View the Project on GitHub lion-codes/lion-codes

This project consists of GPU research and teaching codes, with specific applications contained in individual directories of the main repository. These applications cover many domains including chemical kinetics, condensed matter physics and quantum chemistry to name several. Key algorithms include Newton-Raphson using GMRES, Lanczos eigen-decomposition, QR and LU decomposition, as well as transforms (eg., Radon) and elementary numerical building blocks like random number generation and integration.

There are a variety of dependencies including Boost, Thrust, CUSP, CUSPARSE, CUBLAS, and OpenCV, as noted in individual makefiles, with paths generally suited to the Lion-GA RCC cluster at Penn State.

We gratefully acknowledge support from the Research Computing and Cyberinfrastructure (RCC) Unit at Penn State, Jason Holmes, hardware donations of Tesla devices from Hewlett Packard and Nvidia, the Nvidia CUDA Research Center Program, as well as computational time available through XSEDE on TACC Stampede. All materials contained herein are offered under the Apache License.

Bill Brouwer + Pierre-Yves Taunay 09/13

"Nul problème ne peut résister aux assauts d'une réflexion soutenue" -Voltaire