Effective SIMD Vectorization for Intel Xeon Phi Coprocessors

Figure 1

SIMD vector compilation infrastructure for function and loop vectorization.