CHIP DESIGNER Nvidia has announced that it will release the source code for its CUDA low level virtual machine (LLVM) based compiler.
Nvidia's CUDA compiler will be able to create code that supports more programming languages and will run on AMD and Intel processors, while previously it ran only on Nvidia's GPUs. The company made the announcement today at the GPU Technology Conference in Beijing.
Sudhakar Yalamanchili, professor at the Georgia Institute of Technology said, "Opening up the CUDA platform is a significant step. The future of computing is heterogeneous, and the CUDA programming model provides a powerful way to maximize performance on many different types of processors, including AMD GPUs and Intel x86 CPUs."
The firm believes that releasing the code will accelerate the development of exascale computing by enabling alternative approaches to programming. Essentially the CUDA compiler can be used like Java, creating a virtual machine that runs above the underlying hardware architecture.
Nvidia said, "By releasing the source code to the CUDA compiler and internal representation (IR) format, NVIDIA is enabling researchers with more flexibility to map the CUDA programming model to other architectures, and furthering development of next-generation higher performance computing platforms."
The source code of the CUDA compiler will be available to qualified academic researchers and software tools developers by registering with Nvidia on an early access basis. µ