Siirry suoraan sisältöön

Multicore Programming with Java (3 cr)

Code: TX00EL43-3001

General information


Enrollment

04.05.2020 - 02.08.2020

Timing

03.08.2020 - 07.08.2020

Number of ECTS credits allocated

3 op

Virtual portion

3 op

Mode of delivery

Distance learning

Unit

ICT ja tuotantotalous

Campus

Karaportti 2

Teaching languages

  • English

Seats

0 - 24

Degree programmes

  • Tieto- ja viestintätekniikan tutkinto-ohjelma
  • Degree Programme in Information Technology

Teachers

  • Jörg Hettel

Groups

  • ICTSUMMER
    ICT Summer School

Objective

Knowledge and understanding:
Today, multicore/multiprocessor hardware platforms are widespread used in all ICT domains, including smartphones, automotive, computer vision, and many others. Multicore programming is the key enablers for performance scaling in compute-intensive applications.

The students will know the concept of multicore programming with Java. They will know the concurrency related problems, like race conditions and visibility constraints and their consequences. They have an understanding how these problems can be avoided or solved. They know low-level and high-level synchronization concepts and their application.

Further they know the modern parallelization frameworks of Java, like ForkJoin, parallel Streams or CompletableFuture and understand their internal working mechanisms.

Skills:
The students are able to design and write multicore programs and apply concepts, like synchronization and threadpools. They can design and implement solutions like data or task parallelism. For concrete problems they can decide which parallelization concepts are best suited. They can apply parallelization frameworks to speed up program execution.

Content

Introduction

Concept of Threads

Low-level synchronization concepts
- Synchronized, wait, notify, notifyAll, Locks, Atomics

Java Memory Model

Executors and Futures

High-level synchronization concepts
- CountDownLatch, CyclicBarrier, Phaser

Concurrent Data Structures

Concurrency Patterns

Frameworks for Parallelisation
- Fork-Join-Framework
- Parallel Streams (data parallelism)
- CompletableFuture (task parallelism)
- Reactive Streams (event parallelism)

Case Studies

Evaluation scale

0-5

Assessment criteria, approved/failed

Students actively work on the daily exercises assigned on the course. At the end of the course the examination is divided into two parts: a short written exam (30 min) and practical programming exercises (90 min).

Qualifications

Good Java skills. Basic mathematics, especially vector and matrix operations.

Further information

Intensive course with hands-on exercises on paper and with the computer. A computer with a Java 11 (or higher) development environment is required.