AMD has renamed its wellknown OpenCL SDK, called ATI Stream SDK, in AMD APP SDK. APP stands for Accelerated Parallel Processing.
What is the AMD APP Software Development Kit?
The AMD APP Software Development Kit (SDK) is a complete development platform created by AMD to allow you to quickly and easily develop applications accelerated by AMD APP technology. The SDK allows you to develop your applications in a high-level language, OpenCL™ (Open Computing Language).
More details on the official page: AMD APP SDK hompage.
If you want to test the SDK, you can download it from this page: AMD APP SDK Downloads.
But if you already have ATI Stream SDK v2.3, don’t waste your time: the current AMD APP SDK file is still the same file than ATI Stream SDK 2.3.
6 thoughts on “[OpenCL] ATI Stream SDK Becomes AMD APP SDK”
First a rebranding of the 67XX and now a name change. AMD is all marketing omg!
Okay rebranding cards is not cool, but in the case of the OpenCL SDK, I think it’s for standardizing their tools with the name “AMD”.
AMD is definetly turning in to a marketing troll, and they really need to push on R&D expenditures.
Their current architecture is still way too much graphics-centric and lagging behind GPGPU trends. The other troublesome side is the lack of broad software support and aggressive ISV relations.
Its not as bad as NVIDIA trawling the archives to come up with ‘Ti’
@ — Couldn’t agree more. It’s frightening how little they care about supporting the GPGPU ecosystem.
If they don’t change their role in the market soon (< 1yr), they will have no chance of catching up to NVIDIA.
“Their current architecture is still way too much graphics-centric and lagging behind GPGPU trends”
GPGPU trends tends to differ from GPU trends.
GPU for GPGPU is overloaded with mem enhancements and thread execution enhancements which are of little or no use in rendering. GPU for graphics is overloaded with TMUs, ROPs, UVD/PV, tesselators etc. which are useless in computing. There’s no real future for universal units for BOTH purposes. Either ROPs etc. becomes software @ processing units and not fixed units (which is quite impossible right now) or there should be compute element on one die connected to GPU specific elements on the other (so GFX cards would have both, while compute cards only SIMD Arrays/Stream Multiprocessors with MEM interface).
Comments are closed.