2. EPCC’s MSc Programmes
• On-campus full-time (1 year) and part-time (2-3 years) MSc
Programmes in:
• High Performance Computing (HPC): https://edin.ac/2vu0nJn,
• HPC with Data Science: https://edin.ac/2vxqS0y
• From Academic Year 2020/21 both programmes also available as
online-only 3-6 year part-time intermittent study options.
• Background: Very broad church from computational physicists to
computer scientists, bioinformaticians to statisticians, data scientists
to engineers. Key entry requirements centred around programming.
• Highly practical in focus.
3. MSc content
• Core skillsets taught on programme:
• Parallel programming using OpenMP and MPI
• Good software development practices and project management
• Utilising parallel architectures (compulsory HPC, optional HPCwDS)
• Data Management, Data Analytics and Big Data (compulsory HPCwDS, optional HPC)
• Extension topics and areas (optional courses):
• GPU programming (incl. CUDA, OpenCL, OpenACC), Modern C++, PGAS
• Hybrid Parallel Models, Parallel I/O, Advanced MPI
• Performance and optimisation of code
• Design of parallelisation strategies
• Algorithmic foundations for parallelisation
• Cloud and Distributed systems
• *Neural Nets and Machine Learning
• *Probabilistic Modelling and Reasoning
• *Text processing, retrieval and classification
• *Advanced utilisation of database management systems, data structures, storage
architectures
*= HPC with Data Science only
4. MSc Industrial Projects
• MSc programme at EPCC has run for over
18 years.
• Industrial collaborations have existed for
much of that.
• Projects have been undertaken with
companies ranging from multinationals to
start-ups.
• Projects are able to be bespoke to suit the
individual situation.
• Many previous industrial collaborators have
returned year after year.
• Some have gone on to hire/sponsor their
project students.
Recent partners include:
5. MSc Industrial Projects: the Basics
• Dissertation projects based on real-world problems
• Undertaken between mid-late May-August
• Project Preparation, including planning, takes place January-
March/April
• Student benefits: experiential learning
• Company benefits:
• Cost-effective way to undertake research projects/investigate possible
areas for future development
• Access to EPCC facilities and expertise
• Potential to develop output further through ongoing work with EPCC
• Further information available here: https://edin.ac/2vxh5Yg
6. Project Proposals & Selection: When,
What, How?
• Project proposals (even draft) recommended to be submitted
by 15 October (projects released to students), but no later than
01 November.
• Opportunities to present projects to students on Thursdays 17,
24, 31 October.
• Pre-selection interview/application process recommended to be
undertaken (if wanted) in late October/early November
• Students submit choices mid-November.
• Students assigned to projects and supervisors assigned in
December
7. Project Proposals: The Basics
• Project background - Give a description setting the context to your
project.
• Project proposal: State the aims of your project.
• Prerequisite skills and experience
• Any prerequisite skills for your project, e.g. you expect levels of proficiency in
C++ or Fortran or Java, etc or a particular MSc optional course.
• Anything that you would expect the student to know before the project starts -
but this can include anything to be worked on during Project Preparation
(generally best to include as a recommended skill).
• You may list these as: Mandatory, Recommended, Desirable
• Will the project require an NDA?
• Will the project require a DPA?
• Is it possible to work on sensitive data at the company only?
• Could the work be done on open-source/synthetic data at EPCC instead?
• Would it be possible to develop on open-source first but test at the company at
the end?
8. So what do I need to do?
• Speak to EPCC colleagues who can assist in formulating
a project.
• No such thing as a bad idea: some may not be suitable,
but some may become so with tweaking.
• You get out what you put in: the more involved industrial
collaborators can be, usually the more useful project
output can be to you.
• Likewise: we will never turn down an interesting
problem/project idea out of hand.
9. Questions?
• There are refreshments and nibbles outside, you will find
me there.
• Contact via:
• E-mail:
• b.morse@epcc.ed.ac.uk
• Industrial.dissertation@epcc.ed.ac.uk
• Telephone:
• 01316 513398
• In person/Postal address:
• Room 2.42 The Bayes Centre, 47 Potterrow, Edinburgh, EH8 9BT