CCES Unicamp

Parallel programming education at the CCES

The gigantic increase in demand for Computing in the last 50 years has allowed the emergence of large-scale computational systems based on computer clusters. These systems have truly revolutionized the lives of societies, enabling everything from complex simulations in areas such as biology and chemistry, to the emergence of computational clouds that made the creation of electronic commerce models possible.

Programming these clusters is not an easy task, and requires special training. The course MC970 (Introduction to Parallel Programming) of the Computation Institute was created by Prof. Guido Araujo to train students in parallel programming models such as OpenMP, CUDA, MPI, etc. In total, the students do 7 practical projects and a final major project.

Given the importance of this area in modern Computer Science education, and its applicability in several areas of knowledge, MC970 has been receiving a growing interest from students of Physics, Chemistry, Biology and Engineering.

In order to support the formation of qualified professionals and scientists in this area, CCES created in 2020 a program to support MC70. In this program, students use the resources of CCES’ Kahuna machine to carry out the practical projects of the discipline, while being exposed to a cluster programming environment.

MC970 trains over 60 students every semester in parallel and distributed programming techniques. In the last two semesters there have been more than 125 students trained.

Related posts

Prof. Claudia Bauzer Medeiros presents a talk on eScience and experiment reproducibility at Texas A&M University

escience

PeNTIOS: a package for Petri Net simulation in Saccharomyces cerevisiae

cces cces

MassCCS: A High-Performance Collision Cross-Section Software for Large Macromolecular Assemblies

cces cces