Publications

Abstract not provided.

The usage and adoption of General Purpose GPUs (GPGPU) in HPC systems is increasing due to the unparalleled performance advantage of the GPUs and the ability to fulfill the ever-increasing demands for floating points operations. While the GPU can offload many of the application parallel computations, the system architecture of a GPU-CPU-InfiniBand server does require the CPU to initiate and manage memory transfers between remote GPUs via the high speed InfiniBand network. In this paper we introduce for the first time a new innovative technology - GPUDirect that enables Tesla GPUs to transfer data via InfiniBand without the involvement of the CPU or buffer copies, hence dramatically reducing the GPU communication time and increasing overall system performance and efficiency. We also explore for the first time the performance benefits of GPUDirect using Amber and LAMMPS applications. © Springer-Verlag 2011.

Abstract not provided.

Abstract not provided.

Abstract not provided.

Abstract not provided.

Abstract not provided.

Abstract not provided.

This document summarizes the work done in our three-year LDRD project titled 'Physics of Intense, High Energy Radiation Effects.' This LDRD is focused on electrical effects of ionizing radiation at high dose-rates. One major thrust throughout the project has been the radiation-induced conductivity (RIC) produced by the ionizing radiation. Another important consideration has been the electrical effect of dose-enhanced radiation. This transient effect can produce an electromagnetic pulse (EMP). The unifying theme of the project has been the dielectric function. This quantity contains much of the physics covered in this project. For example, the work on transient electrical effects in radiation-induced conductivity (RIC) has been a key focus for the work on the EMP effects. This physics in contained in the dielectric function, which can also be expressed as a conductivity. The transient defects created during a radiation event are also contained, in principle. The energy loss lead the hot electrons and holes is given by the stopping power of ionizing radiation. This information is given by the inverse dielectric function. Finally, the short time atomistic phenomena caused by ionizing radiation can also be considered to be contained within the dielectric function. During the LDRD, meetings about the work were held every week. These discussions involved theorists, experimentalists and engineers. These discussions branched out into the work done in other projects. For example, the work on EMP effects had influence on another project focused on such phenomena in gases. Furthermore, the physics of radiation detectors and radiation dosimeters was often discussed, and these discussions had impact on related projects. Some LDRD-related documents are now stored on a sharepoint site (https://sharepoint.sandia.gov/sites/LDRD-REMS/default.aspx). In the remainder of this document the work is described in catergories but there is much overlap between the atomistic calculations, the continuum calculations and the experiments.

Abstract not provided.

Abstract not provided.

Abstract not provided.

Biomolecular simulations have traditionally benefited from increases in the processor clock speed and coarse-grain inter-node parallelism on large-scale clusters. With stagnating clock frequencies, the evolutionary path for performance of microprocessors is maintained by virtue of core multiplication. Graphical processing units (GPUs) offer revolutionary performance potential at the cost of increased programming complexity. Furthermore, it has been extremely challenging to effectively utilize heterogeneous resources (host processor and GPU cores) for scientific simulations, as underlying systems, programming models and tools are continually evolving. In this paper, we present a parametric study demonstrating approaches to exploit resources of heterogeneous systems to reduce time-to-solution of a production-level application for biological simulations. By overlapping and pipelining computation and communication, we observe up to 10-fold application acceleration in multi-core and multi-GPU environments illustrating significant performance improvements over code acceleration approaches, where the host-to-accelerator ratio is static, and is constrained by a given algorithmic implementation. © 2010 IEEE.

Abstract not provided.

Abstract not provided.

Abstract not provided.

Abstract not provided.

Abstract not provided.

Abstract not provided.

LAMMPS is a classical molecular dynamics code, and an acronym for Large-scale Atomic/Molecular Massively Parallel Simulator. LAMMPS has potentials for soft materials (biomolecules, polymers) and solid-state materials (metals, semiconductors) and coarse-grained or mesoscopic systems. It can be used to model atoms or, more generically, as a parallel particle simulator at the atomic, meso, or continuum scale. LAMMPS runs on single processors or in parallel using message-passing techniques and a spatial-decomposition of the simulation domain. The code is designed to be easy to modify or extend with new functionality.

Abstract not provided.

We have performed molecular dynamics simulations of cascade damage in the gadolinium pyrochlore Gd{sub 2}Zr{sub 2}O{sub 7}, comparing results obtained from traditional methodologies that ignore the effect of electron-ion interactions with a 'two-temperature model' in which the electronic subsystem is modeled using a diffusion equation to determine the electronic temperature. We find that the electron-ion interaction friction coefficient {gamma}{sub p} is a significant parameter in determining the behavior of the system following the formation of the primary knock-on atom (here, a U{sup 3+} ion). The mean final U{sup 3+} displacement and the number of defect atoms formed is shown to decrease uniformly with increasing {gamma}{sub p}; however, other properties, such as the final equilibrium temperature and the oxygen-oxygen radial distribution function show a more complicated dependence on {gamma}{sub p}.

Abstract not provided.

Application performance is determined by a combination of many choices: hardware platform, runtime environment, languages and compilers used, algorithm choice and implementation, and more. In this complicated environment, we find that the use of mini-applications - small self-contained proxies for real applications - is an excellent approach for rapidly exploring the parameter space of all these choices. Furthermore, use of mini-applications enriches the interaction between application, library and computer system developers by providing explicit functioning software and concrete performance results that lead to detailed, focused discussions of design trade-offs, algorithm choices and runtime performance issues. In this paper we discuss a collection of mini-applications and demonstrate how we use them to analyze and improve application performance on new and future computer platforms.

Abstract not provided.