Identifying Data					2022/23
Subject (*)	Parallel Processing			Code	614G02023
Study programme	Grao en Ciencia e Enxeñaría de Datos
Descriptors	Cycle	Period	Year	Type	Credits
	Graduate	1st four-month period	Third	Obligatory	6
Language	Spanish
Teaching method	Face-to-face
Prerequisites
Department	Enxeñaría de Computadores
Coordinador	Enes Álvarez, Jonatan		E-mail	jonatan.enes@udc.es
Lecturers	Enes Álvarez, Jonatan		E-mail	jonatan.enes@udc.es
Web
General description	Nesta materia o alumno estudará o papel que o uso do paralelismo ten á hora de acelerar a execución de programas en xeral, e o procesamento de datos en particular. O coñecemento teórico partirá dos conceptos técnicos máis básicos de paralelismo, incluíndo a súa utilidade, aplicabilidade e o contexto técnico da execución de programas paralelos (Tema 1). Despois expoñeranse as principais tecnoloxías hardware que existen actualmente para o procesamento paralelo, así como as súas capacidades técnicas subxacentes relacionadas e necesarias para explotar o paralelismo (Tema 2). A continuación, profundizarase no paralelismo con conceptos engadidos, clasificacións, posibles deseños para a súa implementación software e maneiras de analizar o seu rendemento (Tema 3). Finalmente, aplicarase todo o coñecemento previamente adquirido estudando as ferramentas e tecnoloxías modernas para o procesamento de datos masivos, é dicir, poñendo o foco no Big Data (Tema 4). No aspecto práctico, o alumno realizará diversas sesións cun enfoque incremental a fin de adquirir o coñecemento e a habilidade de programar e/ou despregar solucións de procesamento en paralelo de datos. Empezarase con prácticas dirixidas a adquirir competencias máis técnicas e de 'baixo nivel', e irase progresando cara a solucións máis completas e de máis 'alto nivel'. Estas prácticas realizaranse de forma coordinada coa explicación teórica, de tal forma que se usarán técnicas e tecnoloxías previamente explicadas. De igual forma, as prácticas serán autocontidas e fortemente enfocadas á resolución de problemas ou escenarios concretos. Esta materia ten unha forte dependencia con materias previas como "Fundamentos de Programación I e II", principalmente polo seu requisito técnico para a programación, e "Deseño e Análise de Algoritmos" polo seu coñecemento teórico da análise de complexidade dos algoritmos. En menor medida requírense coñecementos previos da materia de "Fundamentos de Computadores" para comprender o comportamento empírico e o rendemento en xeral dalgúns programas cando estes execútanse nun computador.

Study programme competencies Code Study programme competences A12 CE12 - Capacidade de coñecer e aplicar os principios fundamentais, principais paradigmas e técnicas da programación paralela e distribuída ao desenvolvemento de algoritmos para o procesamento e análise masiva de datos. B2 CB2 - Que os estudantes saiban aplicar os seus coñecementos ao seu traballo ou vocación dunha forma profesional e posúan as competencias que adoitan demostrarse por medio da elaboración e defensa de argumentos e a resolución de problemas dentro da súa área de estudo B3 CB3 - Que os estudantes teñan a capacidade de reunir e interpretar datos relevantes (normalmente dentro da súa área de estudo) para emitir xuízos que inclúan unha reflexión sobre temas relevantes de índole social, científica ou ética B4 CB4 - Que os estudantes poidan transmitir información, ideas, problemas e solucións a un público tanto especializado como non especializado B7 CG2 - Elaborar adecuadamente e con certa orixinalidade composicións escritas ou argumentos motivados, redactar plans, proxectos de traballo, artigos científicos e formular hipóteses razoables. B8 CG3 - Ser capaz de manter e estender formulacións teóricas fundadas para permitir a introdución e explotación de tecnoloxías novas e avanzadas no campo. B9 CG4 - Capacidade para abordar con éxito todas as etapas dun proxecto de datos: exploración previa dos datos, preprocesado, análise, visualización e comunicación de resultados. B10 CG5 - Ser capaz de traballar en equipo, especialmente de carácter multidisciplinar, e ser hábiles na xestión do tempo, persoas e toma de decisións. C1 CT1 - Utilizar as ferramentas básicas das tecnoloxías da información e as comunicacións (TIC) necesarias para o exercicio da súa profesión e para a aprendizaxe ao longo da súa vida. C4 CT4 - Valorar a importancia que ten a investigación, a innovación e o desenvolvemento tecnolóxico no avance socioeconómico e cultural da sociedade.

Learning aims

Learning outcomes	Study programme competences
Know of and understand the technical requirements and the current technologies that allow for parallelism.	A12	B8 B9
Know of the different currently available technologies to implement parallelism, their applicability, limits, advantages and disadvantages.	A12	B2 B4 B8 B9
Be able to use parallelism techniques to adapt existing solutions so that they allow parallel processing.	A12	B2 B4 B7 B8 B9 B10	C1
Be able to analyze the performance if a processing solution, with and without parallelization.	A12	B2 B4 B7 B8 B9 B10	C1
Understand the paper that parallelization plays in today's society when it comes to key data processing tasks in business and research.	A12	B3 B4 B8 B10	C4

Contents

Topic	Sub-topic
Chapter 1 - Introduction and previous concepts	* The process and sequential program * Lifecycle of a process * Threads * Paralell program * Usefulness of parallelism
Chapter 2 - Hardware parellelism, hierarchy	* Levels of parallelism * Internal processor parallelism (hidden) * Processor functionalities (low-level parallelism) * Processor accessible resources (high-level parallelism) * Pool of machines (Cluster and Supercomputer) * Distributed computing * Specific devices * State of the art of processors
Chapter 3 - Software parallelism, design and implementation	* Flynn taxonomy * Frameworks and languages for parallelism * Key concepts * Paradigms for parallel processing * Parallel programs analysis * Parallel programs design
Chapter 4 - Big Data technologies	* Data storage * Resource and execution management * Batch processing * Streaming processing

Planning

Methodologies / tests	Competencies	Ordinary class hours	Student’s personal work hours	Total hours
Guest lecture / keynote speech	A12 B3 B8 B9 C4	20	30	50
Laboratory practice	A12 B2 B4 B7 B9 B10 C1	20	70	90
Objective test	A12 B2 B4 B7 B9 C1 C4	3	1	4
Personalized attention		6	0	6
(*)The information in the planning table is for guidance only and does not take into account the heterogeneity of the students.

Methodologies

Methodologies	Description
Guest lecture / keynote speech	* Theory sessions will introduce the basic knowledge later used on practice sessions. * Other concepts will also be explained in detail, either because they are key to understand the technologies and techniques used on the practice sessions, or because they are more advanced and are crucial to understand the paper that parallelism has on nowadays society. ________________________________________________________________________________
Laboratory practice	* Practice sessions will be self-contained and will deal with several specific problems or scenarios where parallelism plays an important role and where previously explained techniques or technologies are used. * Each practice will focus on a single scenario or problem and will be composed of previous description and explanation, a proposed code to be analyzed and used, and a series of questions to work on. The student will have to work on the practice, starting on its first practice session and then continuing on its out-of-classroom time. The questions can range from performing an extension of the code, to performing an empirical study of its performance using several parallelism configurations, describing its behavior or functioning, or other types of questions overall focused at assessing the degree to which the student comprehended the problem and the solution. * It is possible that for some practices, a brief quiz will be used. Nevertheless, such quiz will only be carried out once the practice has finished and submitted by all the students. ________________________________________________________________________________
Objective test	* At the end of the term, and exam will be carried out to evaluate all the subject's knowledge, primarily the concepts from the theory sessions, but also to a lesser extent the ones from the practice sessions.

Personalized attention

Methodologies Guest lecture / keynote speech Laboratory practice

Description * Personalized attention will focus on supporting the students with the overall subject. * On the one hand, personalized attention will be available for those that have some issue understanding any concept exposed on the theory sessions, so that no student has any difficulty in keeping up with the classes and with those topics that will be the subject of evaluation. * On the other hand, personalized attention will also be available for any student that requires some help with specific issues that arise from the practice lessons, whether they are due to technical problems or more deep understanding issues of the key concepts dealt with. Although this help will be available for any practice lesson throughout the term, it is advisable to deal with any doubt or problem either during the practice lesson or shortly afterwards. ____________________________________________________________________________ Those students with an approved dispensation for non-attendance at classes can also benefit by using this personalized attention to ask for the practice briefing as it was given during the ordinary practice classes.

Assessment

Methodologies	Competencies	Description	Qualification
Laboratory practice	A12 B2 B4 B7 B9 B10 C1	* All the practice lessons will be the subject of evaluation and assessment by the teacher. * Each practice lesson will be introduced and briefly explained by the teacher on its first associated practice class. The student is expected to start the practice lesson right away. * The submission deadline of practice lessons will be previously agreed on, a time during which it is expected that the student carries out such practice lesson during the out-of-class time. The deadline will be group-specific. * It is possible that for some practice lessons, the assessment score will be based partially or totally on a quiz that will be carried out on a date known beforehand.	50
Objective test	A12 B2 B4 B7 B9 C1 C4	* Written exam carried out at the end of the term. * It will mainly evaluate and assess concepts from the theory lessons. * To a lesser point, some questions will also be present to re-asses key concepts from the practice lessons.	50
Assessment comments
In order to pass the subject, a minimum of 40% is required on the objective test, or final exam (2 points out of 5). Practice sessions will be NON REPEATABLE for the second chance. Part-time students can attend any practice class group, once it has been previously notified. Part-time students or students with approved dispensation for non-attendance at classes can submit their practice lessons taking into account the longest group-specific deadline available. In case a practice lesson is assessed using a quiz, a different date will be previously negotiated if needed.

Sources of information

Basic	Francisco Almeida et al. (2008). Introducción a la programación paralela. Madrid : Paraninfo Cengage Learning Tomasz Drabas, Denny Lee (2017). Learning PySpark: Build data-intensive applications locally and deploy at scale using the combined powers of Python and Spark 2.0. Packt Publishing Giancarlo Zaccone (2015). Python Parallel Programming Cookbook. Packt Publishing Jesús Carretero Pérez et al. (2007). Sistemas operativos : una visión aplicada . Madrid : McGraw-Hill
Complementary	Peter S.Pacheco (2011). An introduction to parallel programming. Burlington, MA : Morgan Kaufmann Bertil Schmidt et al. (2017). Parallel programming : concepts and practic. Cambridge, MA : Morgan Kaufmann  Wes McKinney (2011). Python for Data Analysis: Data Wrangling with Pandas, NumPy, and IPython. O'Reilly

Recommendations

Subjects that it is recommended to have taken before

Design and Analysis of Algorithms/614G02011 Fundamentals of Computers/614G02005 Fundamentals of Programming II/614G02009 Fundamentals of Programming I/614G02004

Subjects that are recommended to be taken simultaneously

Subjects that continue the syllabus

Advanced Parallel Processing /614G02034

Other comments

It is recommended to have some knowledge and ability to program with Python, as all it will be the language used for all of the practice lessons. It is recommended to have some degree of expertise with a Linux operating system, mainly process and filesystem management.

(*)The teaching guide is the document in which the URV publishes the information about all its courses. It is a public document and cannot be modified. Only in exceptional cases can it be revised by the competent agent or duly revised so that it is in line with current legislation.