Â鶹´«Ã½AV

• Dr Irene Moulitsas

The module aims to provide an understanding of a variety of computational methods for integration, solution of differential equations and solution of linear systems of equations.

The module explores numerical integration methods; the numerical solution of differential equations using finite difference approximations including formulation, accuracy and stability; matrices and types of linear systems, direct elimination methods, conditioning and stability of solutions, iterative methods for the solution of linear systems. Several of the studied numerical methods are implemented from scratch during the lab sessions, and the theoretical properties are then empirically studied and understood.

On successful completion of this module you should be able to:

• Formulate and assess numerical integration methods,
• Use appropriate techniques to formulate numerical solutions to differential equations,
• Evaluate properties of numerical methods, and choose the appropriate method to implement towards the solution of a given differential equation,
• Evaluate properties of systems of linear equations, and choose the appropriate method to implement towards the solution of a given system of linear equations,
• Assess the behaviour of the numerical methods and the computed numerical solutions.

• Dr Irene Moulitsas

Object oriented programming (OOP) is the standard programming methodology used in nearly all fields of major software construction today, including engineering and science and C++ is one of the most heavily employed languages. This module aims to answer the question ‘what is OOP’ and to provide the student with the understanding and skills necessary to write well designed and robust OO programs in C++. Students will learn how to write C++ code that solves problems in the field of computational engineering, particularly focusing on techniques for constructing and solving linear systems and differential equations. Hands-on programming sessions and assignment series of exercises form an essential part of the course. The library support provided for writing C++ programs using a functional programming approach will also be covered.   

An introduction to the Python language is also provided.

• The OOP methodology and method, Classes, abstraction and encapsulation,
• Destructors and memory management, Function and operator overloading, Inheritance and aggregation, Polymorphism and virtual functions, Stream input and output,
• Templates, Exception handling, The C++ Standard Library and STL.

Functional programming in C++.

On successful completion of this module you should be able to:

• Apply the principles of the object oriented programming methodology - abstraction, encapsulation, inheritance and aggregation - when writing C++ programs,
• Design and create robust C++ programs of simple to moderate complexity given a suitable specification,
• Construct C++ programs to solve a range of numerical problems in computational engineering using both OO and functional based approaches, based on the Standard Template Library and other third-party class libraries,
• Design development environments and associated software engineering tools to assist in the construction of robust C++ programs,
• Evaluate existing C++ programs and assess their adherence to good OOP principles and practice.

This course aims to provide a more in-depth look at the software life-cycle phases, requirements engineering and design of applications.

• Software Development Life Cycles
• Requirements Analysis (user requirements, systems requirement specification, functional & non-functional requirement, software requirements specification, modelling and prototyping, structured analysis, formal specification)
• System modelling: UML diagrams, behavioural models, structural models, introduction to model checking and formal specifications
• Design & Implementation: design fundamentals, design architecture, architectural styles, design patterns, data design, objected oriented design, distributed systems design, component based design, data flow oriented design, data oriented design, real time design

This course aims to address issues concerned with related to the quality of software, validation and verification, and software standards.

• Verification & Validation (verification and validation planning, software testing techniques & strategies, preparing test cases and test environments, unit testing and mocks/fakes/stubs, integration testing, validation testing, system testing, regression testing, customer acceptance testing, advanced debugging, automated static analysis, test documentation, Test Driven Development)
• Test Management.
• Software Quality (Introduction to Software Quality, Quality Management (Assurance, Planning and Control), Product quality and process quality, Quality Metrics and Standards, Process Improvement, Reliability).
• Evolution & Software Maintenance (version & change control, maintainability & maintenance, re-engineering, refactoring, reverse engineering).
• Software Engineering Standards (ISO, IEEE).

On successful completion of this module you should be able to:

Design and implement software testing strategies, including documentation of testing requirements and outcomes at each stage of the software design and implementation process; 

Appraise the importance of software testing within the context of the software development lifecycle; 

Evaluate the relationship between software testing and software quality; 

Assess the chief factors influencing the evolution of software, including the need for version control, documentation and planned maintenance tasks;  

Identify the key standards relevant to software quality and assess their impact on each phase of the development lifecycle.

The aim of this module is to teach you the modern computational skills on a key HPC cluster platform. Many interesting scientific problems require analysis or solution of large data sets. For such problems, harnessing distributed computing and storage resources is clearly of great value. Furthermore, the natural parallelism inherent in many analysis procedures makes it feasible to use distributed resources efficiently. 

• The focus of this module is on parallel algorithms and domain decomposition techniques which are suitable for simulation on High Performance Distributed Computing systems. Emphasis is on algorithms for execution on loosely coupled distributed systems, like grid-systems.
• Data-intensive computing algorithms like distributed data mining and data warehousing.
• Parallel numerical algorithms to solve model applications will be discussed and studied through implementation in the hands-on part of the course.
• Load-balancing methods and domain decomposition techniques will be introduced.

On successful completion of this module you should be able to:

Appraise the need to carry out scientific computing on the HPC platforms and demonstrate relevant critical awareness. 

Formulate elements of parallel program design. 

Relate with the challenges and limitations of the parallelisation process including scalability (Amdahl and Gustafson laws) and load-balancing.  

Create a program to solve a practical problem in scientific computing on the HPC platforms by applying knowledge of intensive computational parallel/distributed techniques. 

Assess the performance and efficiency of a parallel program and propose appropriate code improvements and employ techniques to validate and verify the results of parallel programs.

The advent of multi-core processors in the commodity desktop computer market has shifted the emphasis from traditional single threaded computing models to more advanced methods in order to take advantage of the additional processing power that is now available. This has implications for both the traditional high performance computing sector and the workstation market. This course aims to explore the different parallel processing techniques now available on small scale computer systems, such as multi-core desktop computers and GPU devices.

Introduction to Parallel and Multi-Threaded Programming, Safety and Liveness: Synchronisation Techniques, OpenMP – concepts, structures and usage.Using CUDA to solve general purpose problems on the GPU, Software Tools (debugging and optimisation).

Computer graphics is a key element in the effective presentation and manipulation of data in engineering software.  The aim of this module is to provide an in depth practical understanding of the mathematical and software principles behind 2D and 3D visualisation using the widely used OpenGL (desktop) and WebGL (web based) graphic libraries. Representative GUI based 2D and 3D OpenGL/WebGL applications using both Javascript/HTML5 and the Qt development environment are employed. The module will also cover some of the more advanced rendering techniques including lighting, texturing and other image mapping methods used to enhance visual interpretation of data. An introduction to the implementation and use of Virtual Reality in engineering completes the module. Hands-on exercises and an assignment supplement the learning process.

• Mathematical principles behind 2D and 3D visualisation, The graphic and coordinate pipelines, Matrix transformations, Modelling, viewing and projection, OpenGL and WebGL libraries, GLSL shader programming.for the graphic pipeline and GPU
• Development of interactive CG applications using OpenGL, WebGL, GLSL and Qt
• Advanced rendering techniques, lighting, texturing and image mapping
• Introduction to virtual reality.

On successful completion of this module you should be able to:

• Apply the principles underlying the graphic and coordinate pipelines to display and manipulate 2D and 3D models.
• Use the mathematical basis behind 2D/3D modelling and viewing to solve visualisation problems in OpenGL and WebGL.
• Understand, implement and use GLSL shader programs for implementing the graphic pipeline.
• Create interactive visualisation applications using OpenGL/ WebGL, GLSL and Qt.
• Evaluate the use of VR and other advanced technologies for engineering visualisation.

• Dr Antonios Antoniadis

This module aims to provide you with skills in the areas of: the software quality and project management; technical/engineering applications, and cluster or cloud computing so that you can undertake a group project.

• Project Management (risk analysis, estimation models, project planning and scope definition, communication and team working).

• Software Metrics & Quality Assurance (definition, collection, quality metrics, productivity metrics, the review process, software reliability, using software quality metrics).

• Algorithmic stability and performance.

• Specification and performance of computing environments (clusters/cloud, etc).

• Validation and tuning of applications on medium-scale distributed architectures.

On successful completion of this module you should be able to:

1. Demonstrate a systematic understanding of the key management tasks involved in implementing a software project in terms of technical requirements, planning and estimation.
2. Identify and implement the key techniques used in maintaining a software project, such as identification of critical project metrics, comprehensive quality management and risk assessment.
3. Show proficiency in the use of a number of automated tools used to assist the software engineering process, such as project design, documentation and code analysis.
4. Work as part of a team using coordination, negotiation and interfacing skills.

The aim of this module is to provide you with the necessary knowledge and understanding of virtualisation technologies, cloud anatomies and cloud applications to the provision of on-demand computational resources, as well as a wider understanding of how those resources are consumed through Cloud Computing services. 

Cloud Computing Introduction 

Virtualisation 

Cloud Anatomy 

Cloud Security and Economics 

Applied Clouds (AWS, Google Cloud and MS Azure) 

Distributed Systems 

Data Stream Processing 

Applications of Cloud Computing 

On successful completion of this module you should be able to:

1. Demonstrate understanding and knowledge of key virtualisation technologies, cloud anatomies and their applications to Cloud Computing infrastructure;
2. Identify the characteristics of the Cloud Computing platform and understand the fundamentals of distributed systems and how these Cloud characteristics differ from existing distributed computing environments;
3. Critically assess the characteristics of Cloud Computing technologies, and understand the fundamentals of some key computer science applications (e.g., Machine Learning, Big Data) and how these Cloud characteristics affect the implementation of Cloud based applications;
4. Identify the non-technical challenges that affect the implementation and use of Cloud enabled software;
5. Compare and contrast the suitability of different Cloud Computing approaches to different types of computational problem;
6. Develop and implement Cloud enabled software to solve a specified computational problem.