High Performance Resources in Internet Environment

Code:

PRODEI021

Acronym:

REDAI

Keywords
Classification	Keyword
OFICIAL	Comp. Architectures, Operating and Networks Sys.

Instance: 2023/2024 - 2S

Active?	Yes
Responsible unit:	Department of Informatics Engineering
Course/CS Responsible:	Doctoral Program in Informatics Engineering

Cycles of Study/Courses

Acronym	No. of Students	Study Plan	Curricular Years	Credits UCN	Credits ECTS	Contact hours	Total Time
PRODEI	2	Syllabus	1	-	6	28	162

Teaching language

English

Objectives

Objectives:
The main objectives are to provide the students with a solid knowledge of parallel computing (in clusters) and distributed computing in the Internet infrastructure, mainly Grid Computing and Cloud Computing. Also we can enlist as objectives the obtention of a solid knowledge on parallel architectures and on developing parallel programs for distributed memory and shared memory architectures. Through assignments and lab projects the students will also obtain experience in the core technologies in the field, including performance analysis and tuning.

Learning outcomes and competences

Aptitudes and competences:
Algorithm conception and software implementation aiming efficient parallel execution using Foster's methodology.
Use of distribution techniques for grid and cloud environments.
Deployment and exploitation of the developed software in clusters, grids and clouds.

Working method

Presencial

Pre-requirements (prior knowledge) and co-requirements (common knowledge)

General computing.

Proficiency in C/C++ programming for the assignments.

Program

1. Parallel Programming
Introduction to parallel programming, computer architectures, processors, memory organization and interconnection networks.
Parallel Programming Fundamentals: task/channel paradigm, communication patters, synchronization, task granularity and scheduling.
Cluster programming with MPI and OpenMP.
Parallel computing characterization: execution models, programming models, computation models, performance and efficiency measures, scalability analysis.
2. Distributed Computing
Grid computing:
Grid computing models: generic grid, utility grid and desktop grid.
Evolution of grid middleware: metacomputing (e.g. Condor), resource-oriented (e.g. Globus 3) and service-oriented (e.g. Globus 4).
Grid security: authentication, data integrity and encryption, authorization.
Hands-on Grid technology.
Other distributed environments:
Cloud Computing
Peer-to-Peer Computing

Mandatory literature

Quinn, Michael J.; Parallel programming in C with MPI and openMP. ISBN: 007-123265-6
Robert Robey, Yuliana Zamora; Parallel and High Performance Computing, Manning, 2021. ISBN: 978-1617296465
Carminati, Betev, Grigoras; Grid and Cloud Computing: Concepts and Practical Applications, IOS Press, 2016. ISBN: 978-1614996422
Raj, Koleeswaran; Novel Practices and Trends in Grid and Cloud Computing, IGI Global, 2019. ISBN: 978-1522590248
George Reese; Cloud application architectures. ISBN: 978-0-596-15636-7

Teaching methods and learning activities

In the Classes: Theorethical presentation, complemented by examples, small demonstrations and clues for lab assignements and project.
In the Labs: discussions, demonstrantions and problem solving related to assignements and project.

Software

OpenMP
MPI

keywords

Technological sciences > Technology > Computer technology > Software technology

Evaluation Type

Distributed evaluation without final exam

Assessment Components

Designation	Weight (%)
Trabalho escrito	60,00
Trabalho laboratorial	40,00
Total:	100,00

Amount of time allocated to each course unit

Designation	Time (hours)
Estudo autónomo	30,00
Frequência das aulas	42,00
Trabalho de investigação	38,00
Trabalho laboratorial	52,00
Total:	162,00

Eligibility for exams

Implement, deliver and present the assignments.

Calculation formula of final grade

Score = 60% Assignment + 40% Monography

Special assessment (TE, DA, ...)

These students will be subject to all evaluation procedures of regular students, i.e., they must deliver their assignments specified during the course plus any special works also specified. The only difference towards regular students being that they are not required to attend classes, in the cases the law specifically states it.

Classification improvement

The classification improvement can only be done in new instances of the curricular unit.