Publications

Results 26–50 of 97
Skip to search filters

Why Perform Code-to-Code Comparisons: A Vacuum Arc Discharge Simulation Case Study

Contributions to Plasma Physics

Timko, H.; Crozier, Paul C.; Hopkins, Matthew M.; Matyash, K.; Schneider, R.

Numerical modeling is increasingly becoming an indispensable tool for investigations in many fields of physics. Such modeling is especially useful in today's big science projects as a tool that can provide predictions and design parameters. The reliability of simulation results is thus essential. Code-to-code comparisons can help increase our confidence in simulation results, especially when other verification methods - such as comparison to theoretical models or experimental results - are limited or unavailable. In this paper, we describe a code-to-code comparison exercise wherein we compare one-dimensional vacuum arc discharge simulation results from two independent particle-in-cell (PIC) codes. As part of our case study, we define a vacuum arc discharge test problem that can be used by other research groups for further comparison. Early disagreement between the two sets of our results motivated us to re-examine the underlying methods in our codes. After remedying discrepancies, we observe good agreement in vacuum arc discharge time-to-breakdown, as well as in the time evolution of particle and current densities. This exercise demonstrates the usefulness of code-to-code comparisons and provides an example case study for the benefit of other research groups who may wish to carry out similar code-to-code comparisons. © 2012 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim.

More Details

The development of Mellanox/NVIDIA GPUDirect over InfiniBand - A new model for GPU to GPU communications

Computer Science - Research and Development

Shainer, Gilad; Ayoub, Ali; Lui, Pak; Liu, Tong; Kagan, Michael; Trott, Christian R.; Scantlen, Greg; Crozier, Paul C.

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.

More Details
Results 26–50 of 97
Results 26–50 of 97