An alternative method for threading Intel IPP functions is "parallelizing wrappers." This technique is described in "Parallelizing Intel Integrated Performance Primitives Functions Using Intel Cilk Plus and Intel Threading Building Blocks" which can be found in The Parallel Universe magazine issue #6 at http://software.intel.com/en-us/articles/intel-parallel-universe-magazine/.

This technique is especially useful if you want to link against the single threaded version of the Intel IPP library and control threading directly within your application rather than relying on the underlying Intel OpenMP threading that is used in the multi-threaded variants of the Intel IPP library.