OpenVX* is a new standard from Khronos*, offering a set of optimized primitives for low-level image processing and computer vision primitives. The OpenVX specification is applicable across multiple vendors and platforms.
The abstraction of OpenVX notions of resources and execution enables hardware vendors to optimize implementation with a strong focus on a particular platform. At the same time, standardization allows software developers to separate algorithmic innovations from the performance back-ends coming from companies providing accelerators.
For example, the CPU OpenVX implementation from Intel offers graph-level optimizations like tiling and threading, as well as kernel-level efficiency, via intensive use of Intel® Integrated Performance Primitives (Intel® IPP). Similarly, the GPU implementation from Intel relies on the mature OpenCL™ stack to deliver the performance promise.