Since AMD’s OpenCL™ release works on both x86 CPUs as well as AMD GPUs, you can get started coding in OpenCL™ immediately, even if you do not yet have an AMD GPU. In fact, the recommended best practice for programming in OpenCL™ is to start off by writing and debugging your OpenCL™ applications on a compatible x86 CPU. Your OpenCL™ program will be able to automatically scale across all of the available CPU cores on your system. Once your OpenCL™ application is working, you can potentially achieve even more performance gain by adding one or more AMD GPUs and updating your application to take advantage of multiple compatible devices in your system. Get started today by downloading: the latest Accelerated Parallel Processing (APP) SDK release.
All you need to get started accelerating your applications with OpenCL™ is:
- The latest AMD Accelerated Parallel Processing (APP) SDK release
- A supported operating system with supported compiler
- A supported processor (CPU and/or GPU)
OpenCL and the OpenCL logo are trademarks of Apple Inc. used by permission by Khronos.