Our students undertake cutting-edge research to tackle the pervasive parallelism challenge.
The computing industry faces its most disruptive challenge for fifty years. Parallelism has become an essential component of mainstream computer systems, across all scales, from system-on-chip to cloud computing. This pervasive parallelism renders the approaches of the sequential one-step-at-time era increasingly unviable and poses a set of deep research questions, theoretical and practical, pure and applied:
- How should we design parallel programming languages and compilers?
- How should we design and implement parallel architectures and communication networks?
- What theories do we need to prove properties of such systems, or to model and reason about their performance?
- How can concurrent and distributed systems be made secure?
- How can we trade performance for energy in context sensitive ways?
- How can we make algorithms and applications robust against the failures inevitable in exascale systems?
The Centre for Doctoral Training in Pervasive Parallelism addresses this “pervasive parallelism” challenge, educating graduates who will undertake the fundamental research required to transform methods and practices, and become the research leaders of the future. Students develop not only deep expertise in their own specialism, but crucially, an awareness of its relationships to other facets of the challenge. This cross-cutting approach enables us to unlock the true potential of current and future technologies. Our industrial partnerships and engagement programme ensures that our research is informed by real-world case studies and provides a source of diverse internship opportunities for our students.