A pan-European project has started this month to bring together the technologies needed for exascale computing, tackling the key challenge of power usage. The project started this month, bringing together three existing exascale projects on FPGA accelerators, interconnect and 3D chip technologies to reach performance of 10^18 operations, 10 times that of today's fasest supercomputers. At the University of Manchester they are working on OpenCL as the programming model to configure modules that can be plugged into a system as an HPC accelerator.
The final project at University of Pennsylvania, CIS 565: GPU Programming and Architecture has brought us Vulkan Forward Plus Renderer. The original idea came from this paper: Forward+: Bringing Deferred Lighting to the Next Level. In this project the students created a Forward Plus (tiled forward) renderer in Vulkan using compute shader to deal with light culling. A lot was learned from Alexander Overvoorde's Vulkan Tutorial and many other places. The project is available on Github along with a complete behind the scenes story, benchmarks and videos.
Imagination Technologies announces a new offering as part of its Imagination University Programme (IUP) that provides the first complete teaching course on mobile graphics. "An Introduction to Mobile Graphics" includes a rich set of teaching materials and practical exercises that leverage Imagination's popular PowerVR graphics processors (GPUs). The new course has been developed so that little to no previous knowledge of graphics is required, and the content is adaptable to fit most teaching methods and structures. This introductory course is based around OpenGL ES 2.0, the most widely deployed and adopted mobile graphics API.
A paper recently published by the Astronomical Society of Australia on how GPGPUs are dramatically changing the landscape of high performance computing in astronomy.This paper identifies and investigates several key decision areas, with a goal of simplifying the early adoption of GPGPU in astronomy. The merits of OpenCL are considered as an open standard in order to reduce risks associated with coding in a native, vendor-specific programming environment, and present a GPU programming philosophy based on using brute force solutions.
The OpenCL University Kit is a set of materials for teaching a full semester course in OpenCL programming. Each lecture includes instructor notes and speaker notes, plus code examples for lectures 2, 3, and 13. A sample application is provided for lecture 9. A number of educational institutions now offer courses in OpenCL programming to help prepare developers for the new era of heterogeneous computing.
The Western Australian Supercomputer Program (WASP) in conjunction with iVEC and the University of Western Australia is currently running an OpenCL course during the 2010 iVEC/WASP summer school programme. The course will concentrate on teaching students the fundamentals of OpenCL with a goal towards helping students gain a fundamental understanding of OpenCL as a technology framework for enabling heterogenous parallel processing. Online slides sets are available as the course unfolds.
Vivante Corporation announced that the Microprocessor Research and Development Center (MPRC) of Peking University (Beijing, China) selected Vivante 2D/3D GPU IP technology for its Single-Chip-PC SoC Platform, placing the graphics power of OpenGL ES 2.0, OpenGL ES 1.1, and OpenVG in an affordable, 3C computer for schools. "With the Vivante graphics processor and Khronos open standards APIs, we can provide modern graphics applications to benefit education for both teaching and research." said Professor Cheng Xu, Director of the MPRC, Peking University.