high performance computing on graphics processing units: hgpu.org

Unleashing the full potential of heterogeneous systems, consisting of multi-core CPUs and GPUs, is a challenging task due to the difference in processing capabilities, memory availability, and communication latencies of different computational resources. In this paper we propose a novel approach that automatically optimizes task partitioning for different (input) problem sizes and different heterogeneous architectures. […]

Graphics processing units have massive parallel processing capabilities, and there is a growing interest in utilizing them for generic computing. One area of interest is computationally heavy computer vision algorithms, such as face detection and recognition. Face detection is used in a variety of applications, for example the autofocus on cameras, face and emotion recognition, […]

Heterogeneous computing has become prevalent due to the comput-ing power and low cost of Graphics Processing Units(GPUs). OpenCL provides a programming model where the CPU is the master or host, and compute-intensive portions of an algorithm are offloaded to the GPU. However, the host-device model is very limiting. In this model, data-dependent, run-time optimizations that […]

Hybrid computational architectures based on the joint power of Central Processing Units and Graphic Processing Units (GPUs) are becoming popular and powerful hardware tools for a wide range of simulations in biology, chemistry, engineering, physics, etc.. In this paper we present a comparison of performance of various GPUs available on market when applied to the […]

Power-hungry Graphics processing unit (GPU) accelerators are ubiquitous in high performance computing data centers today. GPU virtualization frameworks introduce new opportunities for effective management of GPU resources by decoupling them from application execution. However, power management of GPU-enabled server clusters faces significant challenges. The underlying system infrastructure shows complex power consumption characteristics depending on the […]

Fully utilizing the power of modern heterogeneous systems requires judiciously dividing work across all of the available computational devices. Existing approaches for partitioning work require offline training and generate fixed partitions that fail to respond to fluctuations in device performance that occur at run time. We present a novel dynamic approach to work partitioning that […]

Development on embedded devices, even on today’s hardware, limits us to a minimum of third party-library dependencies due to hardware memory and power restrictions. In setups requiring intense geometric operations on limited hardware, such as in robotics, this problem can often lead to a tedious reimplementation of matrix, vector, and quaternion operations. Furthermore, certain unnecessary […]

We address the low popularity of C++ in computational science by introducing a set of orthogonal libraries: The CUDA-, OpenCL-, and OpenMP-enabled linear algebra library ViennaCL, the mesh datastructure library ViennaGrid, a data storage facility named ViennaData, and the symbolic math kernel ViennaMath. Finally, we discuss how these orthogonal components interact within the finite element […]

Graphics Processing Units (GPUs) are increasingly used for general-purpose applications because of their low price, energy efficiency and enormous computing power. Considering the importance of GPU applications, it is vital that the behaviour of GPU programs can be specified and proven correct formally. This paper presents our ideas how to verify GPU programs written in […]

Computer based simulation software having a basis in numerical methods play a major role in research in the area of natural and physical sciences. These tools allow scientists to attempt problems that are too large to solve using analytical methods. But even these tools can fail to give solutions due to computational or storage limits. […]

Data analysis is a rising field of interest for computer science research due to the growing amount of information that is digitally available. This increase in data has as direct consequence that any analysis is significantly complex. By using structured representations for the data sets, like graphs, the analysis becomes feasible, but is still time-consuming. […]

In this thesis we present a symbolic execution-based technique for cross-checking programs accelerated using SIMD or OpenCL against an unaccelerated version, as well as a technique for detecting data races in OpenCL programs. Our techniques are implemented in KLEE-CL, a symbolic execution engine based on KLEE that supports symbolic reasoning on the equivalence between expressions […]