The free OpenGL ES CapsViewer app is now available for Android. This tools reads out all hardware capabilities of Android devices and can upload them to an online database that developers can access to check out OpenGL ES 2.0 and OpenGL ES 3.0 capabilities for the various devices out on the market. The database contains OpenGL ES information including extensions, compressed texture formats, capabilities, EGL extensions, device features and available hardware sensors. You can compare up to 8 different devices at once, so developers can easily check out what features are supported by their targeted devices.
The newly-released Kishonti GFXBench 3.0 is comprised of nearly all new tests, including battery, render quality, and the first serious OpenGL ES 3.0 performance metric. Newly introduced is the demanding Manhattan test, utilizing OpenGL ES 3.0-specific complex lighting, particles, and, most important, deferred shading. Tom's Hardware has a complete rundown of GFXBench and the GFXBench website contains lots of results for various mobile devices.
The problem with disruption is that it's so… disrupting. This is also what's so attractive about it. The important question is what's on the intended receiving end of that disruption. If it is your competitors' business models and market positions, OpenCL provides a remarkable lever for achieving those objectives.
CEVA announced that it has enriched its CEVA-CV computer vision real-time library to include more than 750 functions. New functions added in the latest CEVA-CV release include feature detection kernels and object recognition algorithms such as Harris Corner, Hough Transform, Integral Sum, Fast, LBP, SURF, HOG, SVM, and ORB detection and matching. CEVA-CV now also includes kernels required by The Khronos Group's OpenVX 1.0 specification, which is set to become the key standard for cross-platform acceleration of computer vision applications and libraries.
Tom Olson, work group chair of the OpenGL ES API and director of graphics research at ARM has a great blog entry on ASTC texture compression. Not too technical and includes some great examples and lots of links. A great read for the middle of the week.
OpenCL Integrated Performance Primitives (OpenCLIPP) is a library providing processing primitives (image processing primitives in the first version) implemented with OpenCL for fast execution on dedicated computing devices like GPUs. It was designed to be simple to use and to have low overhead. Two interfaces are provided: C Interface similar to the Intel IPP and NVIDIA NPP libraries and C++ Interface. An OpenCL SDK is required to build the library.
Amdahl Software announced the latest major update to its flagship Eclipse plug-in, OpenCL CodeBench. With the recent 1.2 release, developers can now benefit from automatic host-code generation as they are coding their OpenCL kernels.