Edinburgh eScience MSc

1
Edinburgh e-Science MSc

• Bob Mann
• Institute for Astronomy NeSC
• University of Edinburgh

2
Overview of e-Science MSc

• Due to start in September 2005
• 15 students in the first year
• 80 students within five years
• Take broad view of e-Science
• Multi-disciplinary course
• Aiming at suiting a wide range of students
• Build on existing Edinburgh strengths

3
What kind of students?

• We see three types of students
• Science students needing to develop IT skills for
  their future research
• Computer scientists looking to future work in
  scientific computing
• Computer scientists looking to web and Grid
  services in commercial IT
• Challenge to satisfy all three types

4
What kind of course?

• Basic requirements
• Underlying principles and hands-on experience of
  current technologies
• Discipline in software engineering
• Range of options, given range of students
• Requires strength in computer science and in a
  range of application areas

5
What can Edinburgh offer?

• Informatics
• Existing MSc offering a wide range of courses
• NeSC
• Expertise in training in e-science technologies
• EPCC
• HPC and mix of CS and application students
• Applications
• Physical, earth and life sciences, plus new
  national Digital Curation Centre

6
Basic programme structure

• Semester 1 (Sep-Dec)
• Four mandatory courses
• Two optional courses
• Semester 2 (Jan-Mar)
• Four mandatory courses
• Two optional courses
• Semester 3 (Apr-Sep)
• Individual research project

Diploma
MSc
7
Mandatory courses

• Semester 1
• Distributed Computing for e-Science 1
• Scientific Software Engineering 1
• Introduction to Scientific Data
• Programming for e-Science
• Semester 2
• Distributed Computing for e-Science 2
• Scientific Software Engineering 2
• Topics in e-Science
• Project Preparation

8
Distributed Computing for e-Science 1 2

• Topics to include
• Models of distributed computation and their
  appropriateness for different tasks
• Generic problems in distributed computing
• Platform dependence, security, fault recovery,
  resource discovery and unavailability
• Hands-on experience of current packages
• Condor, web services, Globus Toolkit, etc
• To be taught by NeSC Training team

9
Scientific Software Engineering 1 2

• Topics to include
• Software development cycle
• From initial requirements to maintenance of code
• Benefits from disciplined software engineering
• Unified Modelling Language (UML)
• Its use in the design of object-oriented software
• Managing distributed software teams
• Development of distributed software
• Developing existing Informatics courses

10
Introduction to Scientific Data

• Topics to include
• The role of data in the scientific process
• Validation, reproducibility, provenance
• Passage from raw state to archived products
• Data management strategies
• Databases, files, XML, etc
• Metadata for data discovery integration
• Case studies from several disciplines
• Seek collaboration with Digital Curation Centre

11
Optional courses

• Three types of optional course
• Courses from existing MSc programmes
• e.g. Informatics, High Performance Computing
• New courses from application areas
• e.g. astronomy - The Virtual Observatory
• Preparing students to undertake projects in
  several Schools within the University
• Advanced (4th/5th year) undergrad. courses
• Wide range of existing options

12
Examples of individual study programmes

• Student wants to work in drug discovery
• Core, plus Bioinformatics 1 2, Applied
  Databases and Learning from Data 1
• Student wants to be astronomer
• Core, plus Virtual Observatory, Astrophysical
  Cosmology, Applied Databases, High Energy Astro.
• Student wants to work in IT industry
• Core, plus Computer Networking, Computer
  Graphics, Parallel Architectures, Parallel
  Algorithms and Programs

13
Current status

• Anticipate first admissions in Sept 2005
• course approved at School College level
• awaiting rubber-stamping at University level
• Seeking funding to support students
• Priority in recent EPSRC CTA bid
• Marie Curie EST proposal in Dec (Dip.?)
• Student Awards Agency for Scotland (Dip.)
• Long-term goal
• All-Scotland (or all-UK?) collaboration