Bulletin of the American Physical Society
APS March Meeting 2014
Volume 59, Number 1
Monday–Friday, March 3–7, 2014; Denver, Colorado
Session M12: Invited Session: The Impact of Heterogeneous High Performance Computing Platforms on Computational Physics |
Hide Abstracts |
Sponsoring Units: DCOMP Chair: Barry Schneider, National Science Foundation Room: 205 |
Wednesday, March 5, 2014 11:15AM - 11:51AM |
M12.00001: Heterogeneous computing. What is it and do we need it? Invited Speaker: Lars Koesterke Heterogeneous computing promises more compute power while consuming less energy. However employing GPUs and Xeon Phi's come with a hefty price tag. Only software that is highly adapted to these new architectures will gain any performance increase. In my talk I will address these questions: \begin{itemize} \item What is heterogeneous computing? \item How can it help producing better results faster? \item Is it worthwhile exploring for the broader audience? \end{itemize} And finally I will try to shed some light on the question whether the future will really be heterogeneous or not. [Preview Abstract] |
Wednesday, March 5, 2014 11:51AM - 12:27PM |
M12.00002: Particle-in-Cell Simulations on Emerging Architectures Invited Speaker: Viktor Decyk Emerging High-Performance Computing (HPC) Architectures consist of multiple accelerators with multiple vector processors. Such platforms require programming with multiple levels of parallelism and pose considerable challenges for performing simulations of plasmas. In this talk we focus on lessons learned from our experience with implementations of Particle-in-Cell (PIC) codes on GPUs. We will discuss aspects of mixed shared memory/distributed memory algorithms, streaming, blocking (tiling), and vector , data structures, latency and load balancing. Many of these lessons are of course familiar from other architectures, but will be seen here in a new perspective. We will discuss strategies for the development of algorithms for PIC codes that are expected to work on a variety of emerging HPC architectures. [Preview Abstract] |
Wednesday, March 5, 2014 12:27PM - 1:03PM |
M12.00003: Using GPUs in Lattice Chromodynamics Invited Speaker: Andrei Alexandru Lattice quantum chromodynamics (QCD) calculations were one of the first applications to demonstrate the potential of GPUs in the area of high-performance computing, because the nature of lattice QCD calculations matches well the GPUs' computational model. In this talk, we will discuss ways to effectively use GPUs for lattice calculations using the overlap operator, a discretization that preserves chiral symmetry even at nonzero lattice spacing and makes possible lattice QCD simulations in the parameter region relevant to Nuclear Physics. We will show that the large memory footprint of these codes requires the use of multiple GPUs in parallel and we will discuss methods used to implement this operator efficiently: mixed-precision for inverters, hybrid CPU/GPU memory use for eigensolvers, and MPI/OpenMP/CUDA parallelization strategies required to take full advantage of both GPU and CPU available resources. We compare the performance of our codes on a GPU cluster and a CPU cluster with similar interconnects. We discuss the strong scaling for problem sizes relevant to current lattice QCD simulations. [Preview Abstract] |
Wednesday, March 5, 2014 1:03PM - 1:39PM |
M12.00004: The Beacon Project: Challenges, Solutions, and Lessons Learned Invited Speaker: R. Glenn Brook With physical limitations imposing increasingly significant performance limitations on future generations of computing hardware, computer architects are turning to increased parallelism and specialized hardware to accelerate key applications and workloads. As a result, emerging high-performance computing (HPC) systems are much more heterogeneous than their predecessors, leading to both operational challenges and application challenges that must be overcome to effectively utilize the associated architectures. With support from the National Science Foundation, the Application Acceleration Center of Excellence (AACE) at the University of Tennessee is currently exploring the impact of the Intel$^{\textregistered}$ Xeon Phi{\texttrademark} coprocessor on computational science and engineering through the Beacon Project, an ongoing research project that encompasses the deployment and operation of an energy-efficient supercomputer and the coordination of an associated research program allowing project teams across the country to explore the applicability of the associated architecture to a variety of scientific codes and libraries. This talk presents an overview of encountered challenges along with associated solutions, highlights some of the current results of the application project teams, and summarizes many of the lessons learned through the Beacon Project to date. [Preview Abstract] |
Wednesday, March 5, 2014 1:39PM - 2:15PM |
M12.00005: Ensembles of AMBER biomolecular simulations on GPUs for assessment and validation of RNA models Invited Speaker: Thomas Cheatham Ensembles of molecular dynamics simulations, using methods including multi-dimensional replica exchange on large-scale GPU clusters such as Blue Waters, provide a fast and efficient means to explore the conformational ensembles of biomolecules such as RNA. Enabling exploration of models in days to weeks instead of months to years we are able to better explore, assess, validate and improve the molecular mechanical force fields for RNA. We will describe our experiences modeling RNA on the large-scale resources and also outline the problems with and improvements in the AMBER suite of programs for simulation and analysis of biomolecules on heterogeneous computing platforms. [Preview Abstract] |
Follow Us |
Engage
Become an APS Member |
My APS
Renew Membership |
Information for |
About APSThe American Physical Society (APS) is a non-profit membership organization working to advance the knowledge of physics. |
© 2024 American Physical Society
| All rights reserved | Terms of Use
| Contact Us
Headquarters
1 Physics Ellipse, College Park, MD 20740-3844
(301) 209-3200
Editorial Office
100 Motor Pkwy, Suite 110, Hauppauge, NY 11788
(631) 591-4000
Office of Public Affairs
529 14th St NW, Suite 1050, Washington, D.C. 20045-2001
(202) 662-8700