Vectorization plays a paramount role in speeding up programs with data parallelism inherent in their algorithms. Intel C++ Compiler can boost performance by detecting the potential parallelism in a loop by checking the data dependence and data flow using its auto-vectorization feature. With contemporary hardware, vector register widths continue to increase and explicit vector programming methods can be used to achieve greater performance. Intel® Cilk™ Plus allows a developer to do explicit vectorization using Intel® Cilk Plus™ SIMD directive, Array Notation and elemental functions to enable the performance potential in your application. Attend this webinar and see it in action, complete with code samples.
Introduction to Vectorization using Intel® Cilk™ Plus Extensions
Pour de plus amples informations sur les optimisations de compilation, consultez notre Avertissement concernant les optimisations.