The OpenCL specification tightly binds a command queue to a specific device. For best performance, the user has to find the ideal queue-device mapping at command queue creation time, an effort that requires a thorough understanding of the underlying device architectures and kernels in the program. In this paper, we propose to add scheduling attributes to the OpenCL context and command queue obj...

OpenCL adoption in the High Performance Computing, entertainment and scientific computing markets continues to grow. The flexibility and portability of OpenCL make it an excellent platform upon which to develop image processing applications. However, OpenCL has not yet been applied to the hardcopy printer and Multi-Function Printer, MFP, markets. The printer/MFP markets traditionally use full c...

FPGA vendors have recently started focusing on OpenCL for FPGAs because of its ability to leverage the parallelism inherent to heterogeneous computing platforms. OpenCL allows programs running on a host computer to launch accelerator kernels which can be compiled at run-time for a specific architecture, thus enabling portability. However, the prohibitive compilation times (specifically the FPGA...

Hardware acceleration using FPGAs has shown orders of magnitude reduction in runtime of computationally-intensive applications in comparison to traditional stand-alone computers [1]. This is possible because on an FPGA many computations can be performed at the same time in a trulyparallel fashion. However, parallel computation at a hardware level requires a great deal of expertise, which limits...

GPU Computing: CUDA vs. OpenCL Currently there is no clear standard for the programming model in applications involving graphics processing units (GPUs). Nvidia as one of the most important hardware manufacturers is pushing their C language extension CUDA, while AMD/ATI as their competitor is following the general OpenCL framework that in principle allows to be applied for arbitrary accelerator...

The use of hardware accelerators in high-performance computing has grown increasingly prevalent, particularly due to the growth of graphics processing units (GPUs) as generalpurpose (GPGPU) accelerators. Much of this growth has been driven by NVIDIA’s CUDA ecosystem for developing GPGPU applications on NVIDIA hardware. However, with the increasing diversity of GPUs (including those from AMD, AR...

In this paper we discuss the interaction of expression templates with OpenCL devices. We show how the expression tree of expression templates can be used to generate problem specific OpenCL kernels. In a second approach we use expression templates to optimize the data transfer between the host and the device which leads to a measurable performance increase in a domain specific language approach...

In this paper AQUAgpusph, a new free SPH software licensed under GPLv3 and accelerated using OpenCL, will be described. Its main differences with respect to other GPU based SPH implementations will be discussed, focusing first on the fact that is accelerated with OpenCL, second on the wide range of solid boundary condition enforcing methods have been implemented (including boundary integrals) a...