Compiler Directives for the Intel Xeon Phi Coprocessor
More InfoLess Info
In this paper, the compiler directives (pragmas) used when writing optimized code for the Intel Xeon Phi coprocessor are introduced and explained. Optimal performance can be realized when key parts of an algorithm are vectorized. This paper goes into detail so the programmer will know the basic approaches to use to transform code so that a compiler can vectorize the code generated in conjunction with the use of compiler directives.