There is a free webinar “Further Vectorization Features of the Intel Compiler” coming next Tuesday talking specifically about getting more vectorizations from Intel Compilers. But you would benefit it more if you've watched/listened to the previous webinar Performance essentials using OpenMP* 4.0 vectorization with C/C++.
Please register at Intel® Software Tools Technical Webinar Series
Detail info on the webinar: Further Vectorization Features of the Intel Compiler
The vectorization features of the Intel compiler continue to get more powerful with each succeeding version. In this webinar, we will look beyond the vectorization of simple loops over intrinsic data types, to examples involving STL (Standard Template Library) vectors; multi-dimensional arrays, including data alignment; explicit outer loop vectorization, using the SIMD feature of OpenMP 4.0; and, if time permits, indirect addressing. Code samples will include C, C++ and Fortran and will be available for subsequent download.