Nvidia broadens Cuda's reach with open-source compiler

Nvidia has opened its Cuda compiler to academics and software vendors to let them port the programming framework to other architectures and languages.The change, announced on Wednesday, means Cuda can be reworked for non-Nvidia GPUs, such as those made by AMD, or even entirely different processor architectures like RISC.

Nvidia has opened its Cuda compiler to academics and software vendors to let them port the programming framework to other architectures and languages.

The change, announced on Wednesday, means Cuda can be reworked for non-Nvidia GPUs, such as those made by AMD, or even entirely different processor architectures like RISC. Cuda is a software framework for programming graphical processing units (GPUs).

"Enabling alternative approaches to programming heterogeneous parallel systems for domain-specific problems and future programming models will help accelerate the path to exascale computing," Nvidia said in a statement. "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."

The source code for the Cuda Low Level Virtual Machine (LLVM) compiler will be released with the Cuda Toolkit 4.1, Nvidia said. Academics and software tool developers can apply for early access to the Cuda compiler source code on Nvidia's website.

Many academics use Cuda to increase the performance of applications running on large supercomputers with many GPUs, such as China's Tianhe-1a or the Mont-Blanc Project's upcoming ARM supercomputer.

Newsletters

You have been successfully signed up. To sign up for more newsletters or to manage your account, visit the Newsletter Subscription Center.
Subscription failed.
See All
See All