Parallel course announcement

ECSE-4740 Applied Parallel Computing for Engineers, Mon & Thurs 4-5:30pm

This engineering course is aimed at seniors and grad students who wish to use current parallel hardware, and to obtain knowledge and hands-on experience in developing parallel applications software. We will use the ECSE parallel computing server, which contains the newest very powerful, parallel hardware: dual 14-core Intel Xeon, Intel Xeon Phi Many Integrated Core (MIC) co-processor, and Nvidia GTX 1080 Pascal graphics card.

Software tools will include CUDA and Thrust on the Nvidia, and OpenMP on the Xeons. Programming paradigms will include multithreaded algorithms using massive (256GB) shared main memory on the Xeons, and single instruction multiple data streams on the Nvidia. Effectively programming these processors also requires in-depth knowledge about parallel programming principles, as well as the parallelism models, communication models, and resource limitations of these processors.

We will see that supercomputers are no longer required for many parallel tasks, and any lab can afford parallel computation.

ECSE-4740 will minimally overlap other RPI parallel programming courses, so that students can profitably take all. Students from outside ECSE are welcome.

Prereq: ECSE-2660 CANOS or equivalent, basic C++.