Code |
14524
|
Year |
1
|
Semester |
S1
|
ECTS Credits |
6
|
Workload |
OT(15H)
|
Scientific area |
Informatics
|
Entry requirements |
--
|
Learning outcomes |
The aim of this course is to study the most important questions and challenges raised when developing high performance applications. It intends to study heterogeneous architectures; big data applications; performance optimization and fault tolerance mechanisms.
|
Syllabus |
1 – Systems for large-scale parallel computing. 2 – Methodologies for developing large-scale parallel programs. 3 – Parallel and distributed parallel algorithms. 4 – Parallel programming languages. 5 – Programming many-core systems (CPU and GPU). 6 – Performance monitoring. 7 –Analysis of causes for performance degradation. 8 – Fault tolerance.
|
Main Bibliography |
High Performance Computing, Modern Systems and Practices. Thomas Sterling, Matthew Anderson and Maciej Brodowicz, 2018, Morgan Kaufmann.
|
Teaching Methodologies and Assessment Criteria |
The teacher will present the concepts that form the state of the art in high performance computing. Through the study and discussion of recent scientific papers in the area, the students will develop skills to analyse and criticise existing systems and eventually propose new solutions.
|
Language |
Portuguese. Tutorial support is available in English.
|