Google has acquired DrawElements, a Finland-based company that specializes in software used for GPU benchmarking and testing. DrawElements' main product is DrawElements Quality Program (dEQP), which is used to measure the quality of GPUs and their drivers, while also focusing on related work with OpenGL ES 2.0 drivers and software rasterizers. The dEQP suite is a toolkit used to benchmark the accuracy, precision, feature conformance and stability of OpenGL ES and OpenCL GPUs.
The refreshed Venue range of Android tablets from Dell is based on Intel Atom processors that run Android 4.4 KitKat. The Atom Z3460/Z3480 SoCs are 64-bit apps processors that include a dual-core CPU based on the Silvermont architecture and a quad-cluster PowerVR G6400 GPU capable of OpenGL ES 3.1, OpenCL 1.2 and RenderScript. In the past OpenCL had only been available on developer boards but the Venue 8 tablet is the first consumer device to come with both OpenCL and RenderScript working out of the box. We’ve loaded our OpenCL Camera Adjustment image processing demo to offer readers a quick comparison in performance. When running on our PowerVR G6400 GPU, the image processing demo averages around 74 fps while the Intel CPU peaks at 12 FPS. To put this into perspective, an Exynos-based Samsung Galaxy S4 using a PowerVR SGX544MP3 GPU was able to run the same demo at roughly 27 FPS.
Boost.Compute is a header-only C++ library for GPGPU and parallel-computing based on OpenCL. This proposed library for Boost is available on GitHub and instructions for getting started can be found in the documentation. Are you interested in Parallel Computing? The Boost.Compute project is currently looking for additional developers. Boost.Compute is not yet an official Boost library.
Going to SIGGRAPH? Be sure to drop into one of our seven Birds of a Feather events, all followed up by our famous BOF Blitz™ After Party. We've got some great news and amazing presentations waiting for you.
Altera has published a new whitepaper written by Acceleware, which compares OpenCL programing on GPUs and Altera FPGAs. The paper provides a brief overview of the OpenCL programming model and then focuses on how OpenCL kernels are executed on Altera FPGAs compared to GPUs. The key differences in optimization techniques for targeting FPGAs is also presented.
Xilinx, Inc recently released the Vivado Design Suite 2014.1, the industry's only SoC-strength development environment. This release extends the automation of the UltraFast design methodology and delivers an average of 25 percent faster runtimes and 5 percent improvement in performance across all devices. Also new to 2014.1 is hardware acceleration of OpenCL kernels, within Vivado High-Level Synthesis (HLS).
Adobe has launched Photoshop CC 2014, a new full-point update that extends the software’s GPU-accelerated Mercury Graphics Engine, and expands its blur effects and Smart Objects functionality. All three new tools are powered by the GPU-accelerated Mercury Graphics Engine, expanded in the update to improve performance on other image-editing tasks: Adobe promises image upsampling is “up to 15 times faster”.
AMD joined other exhibitors at the International Workshop on OpenCL with a unique demonstration. Developed by the AMD FirePro team, the AMD Waterfall demo featured an FTS Celsius R930 Workstation with the AMD FirePro W9100 and a 4K monitor. Designed to showcase OpenCL and OpenGL interoperability, in the waterfall particle simulation, the solution computed the collision of each particle in order to roll it along the rocks. IWOCL 2014 presentations are available for download from Agenda & Slides from 2014 | IWOCL.
FinalWire has announced AIDA64 v4.50. The new AIDA64 release introduces a multi-threaded OpenCL GPGPU AES-256 benchmark to measure cryptographic processing performance of modern graphics processors and APUs.