Probably for reducing the gap with NVIDIA for performance level of its GPU with Open CL environment, AMD decided to release a specific SDK that will allow hardware acceleration for both its GPU and x86 CPU (x86 architecture with SSE3 instruction sets); so it should work on Intel Core CPUs. We currently do not know what will be the performance level of the CPU when compared to the GPU, but it will definitely be interesting to get data and benchmark soon to evaluate how this SDK score vs. NVIDIA's solution.
