Intel® Parallel Computing Center at the LSU Center for Computation & Technology

Principal Investigators:

Honggao Liu, CCT Deputy Director
James A. Lupo, CCT Asst Director for Computational Enablement
Mayank Tyagi, Joint Associate Professor Petroleum Engineering and CCT
Krishnaswamy Nandakumar, Gordon A. and Mary Cain Endowed Chair Professor, Cain Dept. of Chemical Engineering
Karsten Thompson - Chairman, Craft & Hawkins Department of Petroleum Engineering


The Center for Computation & Technology is an interdisciplinary research center at LSU. By uniting experts from diverse fields, ideas are disseminated to foster invention. We concentrate on use of advanced computing infrastructure to conduct research in many different fields which touch disciplines related to science, mathematics, engineering, business, digital media, mass communication, art, music, humanities, and more. The Intel® Parallel Computing Center (Intel® PCC) at CCT will focus on simulation of flows through micropores, such as those found in rocks involved in oil and gas extraction. Current work involving grid generation for porous volumes will be coupled with OpenFOAM to create a new simulation capability with accelerated performance on Intel Xeon Phi™ coprocessor platforms, such as LSU’s SuperMIC HPC cluster.

Flow through porous media is fundamental to many engineering and environmental processes related to oil and gas, chemical processing, hydrology, and geophysics. Simulation challenges are magnified by the complicated geometry of void spaces in rocks or packed bed reactors. Recent advances in HPC capabilities, such as high fidelity mesh generation and flow solvers for these computational domains, have made image-based pore-scale modeling a viable technique for addressing such problems. For example, fines production and migration around a oil wellbore or movement of solid particles in a reactor bed could potentially lead to plugging of pore spaces and subsequent decrease in media permeability. To resolve the underlying physics for realistic applications, multiphase and multiphysics capabilities must be added to current fluid flow solvers. Current community codes, however, do not exploit HPC capabilities to the fullest and lack fully coupled physics. The LSU Intel® PCC will undertake code performance scaling, profiling, and optimization for codes on modern HPC platforms and develop a multiphysics coupled simulation algorithm suited for both fundamental physics needs as well as efficient usage of HPC resources. Discrete phase simulations performed by resolving particle dynamics in a Lagrangian frame will be used along with flow simulations in high-resolution porous media meshes in an Eulerian frame. The computational loads are expected to be high due to large numbers of particles and mesh points, thus close attention will be paid to computational efficiencies. Since open-source community codes will be used, OpenFOAM in particular, the enhancements will become generally available to the user community.

The main barrier to use of accelerators is the steep demand for additional knowledge of languages and methods layered onto those already needed for programming multi-core, multi-CPU nodes. The burden of learning highly complex hybrid programming models presents an enormous software development crisis and demands a better solution. SuperMIC, LSU’s newest 1-PF class Xeon Phi™ coprocessor equipped cluster, will serve as the development platform for extending current programming frameworks, such as OpenFOAM, through incorporation of modules using Xeon Phi™ coprocessor methods. OpenFOAM has an extensive range of features to solve problems ranging from complex fluid flows involving chemical reactions, turbulence and heat transfer, to solid dynamics and electromagnetics. It follows a highly modular code design in which collections of functionality are each compiled into their own shared library. The broader computational sciences community will benefit greatly from the distribution of accelerated modules of OpenFOAM, which may lead to improvements in other CFD related projects.

Related websites:

For more complete information about compiler optimizations, see our Optimization Notice.