This is an archived handbook for a previous year. The current undergraduate and postgraduate handbooks can be viewed using the links above.

Postgraduate Handbook 2018-2019 (archived)

Module COMP52130: Core II B: Advanced Scientific and High Performance Computing

Department: Computer Science

COMP52130: Core II B: Advanced Scientific and High Performance Computing

Type	Level	Credits	Availability	Module Cap	Location
Tied	5	30	Not available in 2018/2019		Durham

Tied To

Pre-requisites

Corequisites

Excluded Combinations

Aims

Provide advanced knowledge and critical understanding of paradigms, fundamental ideas, tools and methods of program performance analysis
Provide advanced knowledge and critical understanding of paradigms, technologies, tools and trends in performance engineering incl skills how to use advanced HPC programming techniques
Provide advanced knowledge and critical understanding of paradigms, fundamental ideas, algorithms and methods behind the modelling and simulation of continuous systems (partial differential equations)
Provide advanced knowledge and critical understanding of paradigms, fundamental ideas, algorithms and methods behind the modelling and simulation of discrete systems

Content

Performance models and analysis
Performance engineering techniques in High-Performance Computing and accelerator programming
Scientific Computing for Partial Differential Equations
Scientific Computing for Discrete Systems (graph-based models, combinatorial optimisation, algorithms and complexity)

Learning Outcomes

Subject-specific Knowledge

understanding and critical reflection of advanced ideas and techniques behind the performance analysis and on performance statements of scientific computing and data analysis codes
understanding and critical reflection of advanced paradigms and relevant techniques in high-performance computing and accelerator programming
understanding and critical reflection of advanced ideas, numerical techniques and algorithms used to study discrete models
understanding and critical reflection of advanced ideas, numerical techniques and algorithms used to study continuous models

Subject-specific Skills

basic familiarity with state-of-the-art software toolboxes to solve large-scale equation systems
competent and educated selection and statements on potential performance of specific codes plus skills to identify shortcomings of existing solutions
competent and educated selection and application of programming languages and analysis tools for specific problems
competence to translate continuous and discrete problem descriptions into algorithmic formulations; competent and educated selection of appropriate solution algorithms

Key Skills

familiarity with advanced paradigms and modern concepts underlying scientific computing both for discrete and continuous systems as well as High-Performance Computing

Modes of Teaching, Learning and Assessment and how these contribute to the learning outcomes of the module

Teaching Methods and Learning Hours

Activity	Number	Frequency	Duration	Total Hours
Lectures in Performance Analysis	4	1 per week	60 minutes	4
Practical Classes in Performance Analysis	12	3 per week	60 minutes	12
Lectures in Performance Engineering	8	2 per week	60 minutes	8
Practical Classes in Performance Engineering	8	2 per week	60 minutes	8
Lectures in Modelling and Simulation of Continuous Systems	12	3 per week	60 minutes	12
Practical Classes in Modelling and Simulation of Continuous Systems	4	1 per week	60 minutes	4
Lectures in Modelling and Simulation of Discrete Systems	12	3 per week	60 minutes	12
Practical Classes in Modelling and Simulation of Discrete Systems	4	1 per week	60 minutes	4
Preparation, reading, and self-study				236
Total				300

Summative Assessment

Component	Component Weighting	Element	Length/Duration	Element Weighting	Resit Opportunity
Coursework	100%	Discrete Systems with Performance Analysis and Engineering	5 weeks	50%
Coursework	100%	Continuous Systems with Performance Analysis and Engineering	5 weeks	50%

Formative Assessment

Formative feedback is given as part of the coursework feedback. The linear algebra workshop is purely formative. Yet, content is used in the lectures.

Students who do not attend monitored activities shown under Teaching Methods and Learning Hours, or who fail to complete the summative or formative assessment(s) specified above, may be subject to the Academic Progress procedures defined in the University's General Regulation V, and may be required to leave the University.