Our colleagues conducted a class together with Intel

publication date: April 3, 2015

On March 31, during the conference PCT'2015 a class on programming for Intel Xeon Phi was conducted by the companies Intel and Singularis Lab. The class was designed for engineers, researchers and professors who work in multicore programming. The class involved a master-class devoted to solving applied tasks with use of Intel Xeon Phi coprocessor with examples of different optimization techniques for recursive matrix multiplication algorithm. A special attention was focused on management with assigning OpenMP threads to the cores, on special features of use Intel MKL library for nested parallelism and on techniques of profiling applications for Intel Xeon Phi.

You can find the class papers in our company web-site, the page http://www.singularis-lab.com/en/papers.html.

Moreover, within the main course of the conference our colleagues Oleg Shapovalov and Victor Getmanskiy presented the reports on Strassen algorithm implementation for Intel Xeon Phi and Optimization of mapping for interconnecting MPI processes to computational architecture. The reports are devoted to the projects that our company implemented together with Intel Inc. The report papers are also available in our web-site, the page Papers.

P1020320web P1020330web