This session will introduce Intel tools, including Intel® Parallel Studio XE and Intel® System Studio, comprehensive suites of development tools to build and modernize code with the latest techniques in vectorization, multithreading, multinode parallelization, memory optimization and an all-in-one, cross-platform tool suite, purpose-built to simplify system bring-up and improve system and IoT device application performance on Intel® platforms.

Step by step, the attendee will be able to understand how to modify the code of an N-body simulation code, simulation of a dynamical system of particles under the influence of gravity forces, thru different optimization stages spanning from scalar tuning, vectorization tuning, memory tuning and threading tuning.

In this lab, we’ll work on the Barnes-Hut tree algorithm to subdivide a family of points in the plane. This algorithm is used in many high-performance computing projects in astrophysics, in the study of the stealth properties of aircraft and the fault openings in the earth crust. From a non-optimized C++ code, we’ll first spend time on vectorization with the help of Intel Advisor before leveraging multithreading with Threading Building Blocks (TBB), that will allow us nested and well-balanced parallelism.