Search

We present a way to improve the performance of the electronic structure Vienna Ab initio Simulation Package (VASP) program. We show that high-performance computers equipped with graphics processing units (GPUs) as accelerators may reduce drastically the computation time when offloading these sections to the graphic chips. The procedure consists of (i) profiling the performance of the code to isolate the time-consuming parts, (ii) rewriting these so that the algorithms become better-suited for the chosen graphic accelerator, and (iii) optimizing memory traffic between the host computer and the GPU accelerator. We chose to accelerate VASP with NVIDIA GPU using CUDA. We compare the GPU and original versions of VASP by evaluating the Davidson and RMM-DIIS algorithms on chemical systems of up to 1100 atoms. In these tests, the total time is reduced by a factor between 3 and 8 when running on n (CPU core + GPU) compared to n CPU cores only, without any accuracy loss. © 2012 Wiley Periodicals, Inc. [ABSTRACT FROM AUTHOR]

Particle tracking methods are central to a wide spectrum of scientific computing applications. To support such applications, this paper presents a compact software architecture that can be used to interface parallel particle tracking software to computational mesh management systems. A detailed description is presented of the in-element particle tracking framework supported by this software architecture--a framework that encompasses most particle tracking applications. The use of this parallel software architecture is illustrated through the implementation of two differential equation solvers, the forward Euler and an implicit trapezoidal method, on a distributed, unstructured. computational mesh. A design goal of this software effort has been to interface to software libraries such as Scalable Unstructured Mesh Algorithms and Applications (SUMAA3d) in addition to application codes (e.g. FEMWATER). This goal of portability is achieved through a software architecture that specifies a lightweight functional interface that maintains the full functionality required by particle-mesh methods. The use of this approach in parallel programming environments written in C and Fortran is demonstrated. [ABSTRACT FROM AUTHOR]