LPGPU.org
An EU-funded research project into low power GPU technology
Low-Power Parallel Computing on GPUs 2 (LPGPU2)” is a EU-funded research project into low powered graphics devices. It is the work of a specially formed consortium of four companies and one university, all from across the EU, who are collaborating to deliver advances in tools and applications for energy efficient use of mobile GPUs.
By continuing to use the site, you agree to the use of cookies. more information
The cookie settings on this website are set to "allow cookies" to give you the best browsing experience possible. If you continue to use this website without changing your cookie settings or you click "Accept" below then you are consenting to this.
LPGPU1 Project
Key Features
Innovation
Massively parallel GPUs are now being used in a great variety of market segments, ranging from video-games, to user interfaces, and to HPC. There are several signs, however, that computer and consumer technology industries are faced with major challenges in delivering improved performance and innovation for future entertainment devices. First, game developers have argued that while GPUs are increasing in performance, this is not leading to visual quality improvements because GPUs fundamentally restrict their flexibility. Second, there are signs that GPUs are approaching a “power wall”, and architecture innovation is required now to circumvent this wall. Third, there is a lack of GPU tools available to compare multi-core processors (CPUs) to GPUs and to perform GPU program transformations to optimize for performance and power. To address these challenges, this project brings together commercial tools, applications and GPU designers, with academic researchers to analyze real-world mass-market software on comparable graphics processor architectures. The project results will help the design of next-generation GPUs, games consoles, and mobile phones, and help software developers produce graphically innovative software in the future. In particular, the project seeks to achieve power and bandwidth reductions of 2x or more on real-world software on next-generation GPUs, as well as GPU architecture designs that are capable of advanced real-time graphics techniques (such as radiosity and game AI) at power levels suitable for battery-powered devices.
Technical approach
The consortium will produce:
The lighting research will start with the Enlighten product from Geomerics, and adapt it to run on a variety of low-power GPUs. We will then analyse Enlighten to see how new improved lighting techniques can be combined with improved GPU architectures to enable more realistic lighting in future games and real-time software.
The Game AI work will enable more interesting interactions with computer-controlled characters in video games, while also enabling GPUs to be able to take on more complex performance-critical tasks than is currently achievable. By working on getting complex C++ game AI code working on GPUS, this will help take GPU computing from scientific research applications into a wider range of applications.
The analysis tools will take existing software, whether written using OpenCL or C++, and analyse memory bandwidth and power usage. These tools help software developers adapt their software for GPUs, while also helping GPU designers make architectural choices.
The analysis and architecture work is all about enabling a next generation graphics processor technology. There are many decisions that have to be made in designing graphics processors, with many hard-to-measure variables. We will produce some real numbers for real-world software, for different architectural choices.
Demonstration and Use
The consortium will produce demonstrations of new lighting and game AI techniques for video games and GPUs. Also, scientific papers will present analysis of GPU software and implications for GPU architecture. These will be presented at prominent graphics and processor architecture conferences. The architecture techniques developed in the project will be integrated in a full system (high-end FPGA) along with a main open-source processor (e.g. open-SPARC LEON-2). By working together, the partners will be able to not only develop the technology, but also test their technologies with those of the partners. This will be particularly helpful in demonstrating the outcomes achieved.
Scientific, Economic and Societal Impact
European companies lead in the design of mobile phone CPUs and GPUs, and are world leaders in video-games technology. These companies need to make large investments in R&D for graphics, for which it is vital that they have reliable information. The main market areas for increased processor performance over the next few years are graphics and video-games where the 4 commercial small technology business companies in the consortium are world leaders in particular in real-time lighting for computer graphics (Geomerics), video game AI (AIGameDev.com), power-efficient GPU design (Think Silicon) and GPU tools (Codeplay). The project is expected to strengthen their market position and a manifold return on investment is envisaged. The universities in the consortium will extend their leading position on low-power computer architecture (Uppsala) and parallel applications and multi-core architectures (TUB).