MAIS

Seminar: Current Topics in Concurrency and Parallelism

Form:	Seminar - 3 CP (2 SWS)
Organizer:	Prof. Dr. Heiko Mantel
Contact:	Florian Dewald
Time and place:	Block seminar, tentative dates: 1.7.2021 and 2.7.2021
Language:	English
Registration:	via TUCaN, course id 20-00-0960-se
Max. participants:	10
Preparation Meeting (online):	Thursday, 15.4.2021, at 15:15
Literature:	You can find relevant articles here. More details will be provided in the preparation meeting.

Information regarding the Corona pandemic:
This course will happen. In the beginning of the summer semester 2021, this course will be taught online. We closely monitor the situation and adapt the format of this course accordingly throughout the semester. The official start of this course is the online preparation meeting on Thursday, 15.4.2021, at 15:15, in which you will receive more information about this course's format. We will provide more information how to join the preparation meeting via e-mail before. To receive this e-mail, please register for this course via TUCaN or write an e-mail to kickoff-2021@mais.informatik.tu-darmstadt.de.

Please register for the preparation meeting by 14.4.2021.

On-line participation in the preparation meeting is required for all labs and seminars. The registration in the courses gets only effective after steps explained in the preparation meeting.

We have distributed the information how to join the preparation meeting. In case you haven't received the credentials yet, please contact us via kickoff-2021@mais.informatik.tu-darmstadt.de as soon as possible.

Content

Picture (modified): Argonne National Laboratory (CC BY-SA 2.0)

Modern computing hardware offers multiple processor cores to complete tasks. Since the performance of a single processor core only increases ever so slightly year over year, more and more software makes use of multiple processor cores. However, developing and analyzing parallel software is a challenging task.

In this seminar, we look at scientific publications that address the creation and analysis of parallel software. The topics include:

dependency analysis and dependency representations,
semantics of concurrent systems,
static and dynamic analysis of concurrent systems,
optimizations for concurrent systems, and
automated parallelization.

Learning Objectives

After successfully participating in this course, you will be able to discuss developments in concurrency and parallelism. Furthermore, you will have improved your skills in reading and understanding scientific articles and in presenting, discussing, and comparing scientific results.

Prerequisites

Knowledge of Computer Science equivalent to the first four semesters in the Computer Science Bachelor program.