CHIPMAKER Intel has released its software development kit (SDK) for the OpenCL Applications XE 2013 integrated development environment (IDE) for Windows and Linux.
Over the past year Intel has warmed to OpenCL as it tries to exploit many-core architectures and its integrated GPUs. The firm announced that the Intel SDK for OpenCL Applications XE 2013 IDE has reached production release status, supporting both Windows and Linux.
Although AMD has been pushing OpenCL for years as a way to make use of the GPU on its accelerated processing unit (APU) chips, Intel has started to embrace the technology with the integrated GPU on its Ivy Bridge processors, supporting full-profile OpenCL 1.2, with Haswell graphics cores doing the same.
Intel’s OpenCL IDE, however, pitches OpenCL as a way to develop applications that make use of many cores, regardless of whether they are vector processors such as GPUs or standard x86 cores such as Xeon or Xeon Phi accelerators.
The Windows version of Intel’s SDK for OpenCL Applications XE 2013 IDE supports the firm’s GPUs along with Xeon and Xeon Phi chips. However Intel’s IDE doesn’t support its GPUs on the Linux version of the IDE, perhaps due to the fact that the supported Linux distributions themselves don’t ship with a Linux kernel that supports the GPUs found in Ivy Bridge processors.
Intel’s growing support for OpenCL could force Nvidia to support the language more rather than push its own CUDA programming language. While Nvidia has enjoyed some success with CUDA, Intel’s blessing of OpenCL should attract even more developers to the programming language.
Intel’s SDK for OpenCL Applications XE 2013 IDE is a free download for Windows, Red Hat Enterprise Linux and Suse Linux.