Center Members

Edmond Chow, Director and Associate Professor

Hua Huang, M.S. student

Ben Pritchard, Post-doctoral Fellow, currently at Virginia Tech

Xing Liu, Ph.D., currently at Intel

Mission

The Intel Parallel Computing Center at Georgia Tech focuses on modernizing the performance and functionality of software for Intel Xeon CPUs and Intel Xeon Phi in the fields of computational chemistry and biochemistry. Both these fields very actively fuel scientific discovery in chemistry, biology, materials science, pharmacology, chemical physics, chemical engineering, and energy research, and are responsible for a significant portion of computer and supercomputer time worldwide. Advances in high-performance applications in these fields promise breakthroughs in solving challenging problems, including real-time and large-scale problems that cannot be solved today.

See the description of the center at the Intel IPCC website.

See announcements of the center from insideHPC and from HPCwire.

Our center was featured in this article in Scientific Computing.

Current Activities

  • Algorithms and software for SIMD vectorized calculation of electron repulsion integrals ubiquitous in quantum chemistry
  • Development of an efficient, distributed memory software framework for computing Coulomb and exchange matrices

Software

SIMINT is a library for calculating electron repulsion integrals. It was designed from scratch to use the SIMD features of Intel Xeon and Intel Xeon Phi CPUs, and is thus highly efficient and faster than other state-of-the-art ERI codes. The approach is to vectorize the computation of primitive integrals of the same type. The Obara-Saika algorithms are used. The developer of SIMINT is Ben Pritchard. SIMINT is automatically generated code. If you are interested, you can also download the code generator.

GTFock is a toolkit for the distributed computation of the Fock matrix in quantum chemistry. It also includes infrastructure for performing self-consistent field (SCF) iterations to solve for the Hartree-Fock approximation. GTFock uses a new distributed algorithm for load balancing and reducing communication and has been demonstrated on 8100 nodes of Tianhe-2. The SCF component of GTFock uses density matrix purification to enhance scalability. GTFock separately constructs generalized Coulomb and exchange matrices can be used to parallelize various methods in quantum chemistry. GTFock can also be used as a benchmark and workload for high-performance computing research.

OptERD is an optimized code for computing electron repulsion integrals, based on the ERD code in the ACES III package. It is about 2-3 times faster than the original code on Intel CPUs and Intel Xeon Phi. The main developers are Xing Liu, Marat Dukhan, and Sanchit Misra.

Hardware

Through an educational grant from Intel, our center operates two Intel Xeon Phi servers, each with 8 Intel Xeon Phi cards. These servers, called "joker" and "gotham," have dual 10-core and dual 16-core Haswell processors, respectively. The servers are used in computing courses at Georgia Tech, for example, CX 4220, Introduction to High Performance Computing.

Protein-ligand system simulated using GTFock.

Publications

B. P. Pritchard and E. Chow, Horizontal Vectorization of Electron Repulsion Integrals, Journal of Computational Chemistry, 37, 2537-2546 (2016). Supporting Information.

E. Chow, X. Liu, S. Misra, M. Dukhan, M. Smelyanskiy, J. R. Hammond, Y. Du, X.-K. Liao, and P. Dubey, Scaling Up Hartree-Fock Calculations on Tianhe-2, International Journal of High Performance Computing Applications, 30, 85-102 (2016). DOI: 10.1177/1094342015592960

E. Chow, X. Liu, M. Smelyanskiy, and J. R. Hammond, Parallel Scalability of Hartree-Fock Calculations, The Journal of Chemical Physics, 142, 104103 (2015). DOI: 10.1063/1.4913961

Other Collaborative Publications with Intel

K. Pamnany, S. Misra, V. Md., X. Liu, E. Chow and S. Aluru, Dtree: Dynamic Task Scheduling at Petascale, ISC High Performance, Frankfurt, Germany, July 12-16, 2015, Lecture Notes in Computer Science, 9137, 122-138 (2015).

X. Liu, M. Smelyanskiy, E. Chow, and P. Dubey, Efficient Sparse Matrix-Vector Multiplication on x86-based Many-core Processors, 27th International Conference on Supercomputing (ICS), Eugene, OR, June 10-14, 2013.

X. Liu, E. Chow, K. Vaidyanathan, and M. Smelyanskiy, Improving the Performance of Dynamical Simulations via Multiple Right-Hand Sides, 26th IEEE International Parallel and Distributed Processing Symposium (IPDPS), Shanghai, China, May 21-25, 2012.

Selected Presentations

“Quantum Chemistry on Large-scale Heterogeneous Clusters,” Sparse Solvers for Exascale: From Building Blocks to Applications, Greifswald, Germany, Mar. 22-26, 2015. (Plenary Talk)

“Partitioning and Load Balancing in Parallel Quantum Chemistry,” Recent Advances in Parallel and High Performance Computing Techniques and Applications, Institute for Mathematical Sciences, National University of Singapore, Jan. 12-16, 2015.

“Scaling Up Hartree-Fock Calculations,” New and Future Directions in Atomistic Simulation and Modeling Workshop, Seattle, WA, Oct 27-29, 2014.

“Computational Methods for Large-Scale Macromolecular Simulations,” Department of Biochemistry Seminar, University of Iowa, Iowa City, IA, Oct. 23, 2014.

“Massive Asynchronous Parallelization of Sparse Matrix Factorizations,” Parallel Computing Lab Seminar, Intel Corporation, Santa Clara, CA, Aug. 14, 2014.

“Large-scale Hartree-Fock Calculations on Heterogeneous Clusters,” Computing Sciences Seminar, Lawrence Berkeley National Laboratory, Berkeley, CA, Aug. 12, 2014.

“Toward Quantum Chemistry on Millions of Cores,” Future of Computational Chemistry Symposium, 248th American Chemical Society National Meeting, San Francisco, CA, Aug. 10-14, 2014.

Collaborators at Intel Parallel Computing Lab

  • Pradeep Dubey, Director and Intel Fellow
  • Mikhail Smelyanskiy, Senior Research Scientist
  • Jeff Hammond
  • Alex Heinecke
  • Sanchit Misra
  • Dheevatsa Mudigere
  • Kiran Pamnany
  • Karthik Vaidyanathan

Education

Teaching materials for high performance computing with an emphasis on scientific computing on Intel processors are available here.