APPLICATIONS

Open Source LLVM Code Generation Project and Emerging OpenCL Standard free RapidMind from Ongoing Development of Device Drivers and Low Level Code Generation: RapidMind, provider of the RapidMind Multi-core Development Platform, today announced that it is both leveraging and contributing to the LLVM (Low Level Virtual Machine) Compiler Infrastructure project as well as participating in and adopting the OpenCL (Computing Language) standard. Until now, RapidMind has dedicated significant resources to the ongoing development and maintenance of low-level, device-specific drivers within the RapidMind platform. Through the use of LLVM and OpenCL, RapidMind can continue to simplify parallel programming for developers, while reducing the investment required in maintaining certain code generation technology. RapidMind will continue to focus on providing an application-oriented runtime platform that is an easy, portable and inherently safe development solution for software organizations needing to take advantage of multi-core processors. Using the RapidMind platform, software organizations are able to leverage multi-core CPUs, GPUs and the Cell Broadband Engine without having to learn new languages or understand the underlying hardware architecture. The software developer does not need to manage complex issues such as load balancing and memory management or rely on inherently risky programming methodologies such as multi-threading. LLVM and OpenCL are just two of the initiatives that RapidMind will be supporting to encourage the growth of a software ecosystem around these emerging processors. RapidMind is augmenting its proprietary code generation technology with LLVM, both to leverage LLVM’s existing code base, and to allow much quicker support for new architectures. When a hardware vendor adds LLVM support for any new architecture, RapidMind can very quickly enable software developers to use RapidMind to build parallel applications for that new processor. LLVM is a set of open-source libraries and tools that provide a hardware-independent interface to generate and optimize machine code for a variety of processor architectures. LLVM is used in a number of compilers and other products that require code generation, such as domain-specific languages and Just-In-Time compilers. According to the LLVM website (llvm.org), companies such as Apple, Adobe, Electronic Arts, Siemens, and others have integrated LLVM into their software. RapidMind is also planning to apply its expertise in multi-core architectures and parallel software development to the project. “As part of our investment in integrating LLVM into the RapidMind platform, we have already contributed a number of improvements back to LLVM and will continue to do so in the future,” said Dr. Michael McCool, RapidMind chief scientist. RapidMind is also a contributor member of The Khronos Group’s working group collaborating to create a new standard, which will service HPC applications through a low-level, high-performance abstraction to form the base layer of an HPC ecosystem. OpenCL, although not yet ratified and still early in its evolution, could help create a more consistent, device-level interface to GPU capabilities that has been lacking. “We are excited about the possibility of a non-proprietary standard for a low-level layer interface that targets GPUs from a spectrum of hardware vendors if they choose to support OpenCL. ISVs using RapidMind today gain the benefits of GPU performance, without having to understand the architecture of the underlying hardware,” said Ray DePaul, president and CEO of RapidMind. “A single standard makes it easier for us to support multiple hardware platforms, and ultimately expand our base of supported hardware.” “The disruption caused by the introduction of multi-core processors, heterogeneous processors, and the increasing adoption of accelerators has created chaos in the software industry and has hindered innovation,” DePaul adds. “There is need for a development and runtime platform for parallel computing. At the foundation of this platform, we require low-level technologies that expose the tremendous potential of the new hardware. Initiatives like LLVM and OpenCL provide the foundation of the ecosystem, allowing RapidMind to create the building blocks needed by software organizations. We are happy to embrace these initiatives.”