Industry-Sponsored Studentships
These projects are sponsored by our industry partners. Students working on these projects will receive funding and supervisory guidance from the sponsoring company, as well as from the CDT.
To apply for one of these studentships, please follow the standard application procedure for the CDT in Pervasive Parallelism, but focus your research proposal on the relevant project described below. You are advised to begin by discussing your application with the named contact for that project.
Optimisation for Deep Learning on Embedded Devices
Sponsored by ARM
Two PhD studentships are currently available under the supervision of Prof. Michael O’Boyle, within the Institute for Computing Systems Architechture in the School of Informatics. The studentships are anticipated in September 2017, but this is flexible.
Both projects are concerned with efficient implementation of deep learning networks on constrained devices. While there has been much activity in how to efficiently learn a network with large training data, there is much less on how to deploy it efficiently on an constrained resource device. The best network and code structure depends on scenario and there will be a trade-off between space, time, energy and accuracy. The projects will investigate code optimisations, such as: code specialisation, higher-parameter exploration, auto-tuned libraries, reduced bit data representation etc to explore these trade-offs. How to update and adapt the network to new data could also be an an area of research.
The exact details of each project are flexible depending on the candidate’s interests and background.
Research Partner
The Institute for Computing Systems Architecture (ICSA) is one of seven research institutes in the School of Informatics at the University of Edinburgh. It was founded in 1998, following the creation of Informatics. ICSA is primarily concerned with the architecture and engineering of future computing systems. Within its five research groups, ICSA covers topics which include: performance and scalability, innovative algorithms, architectures, compilers, languages, and protocols.
The CaRD group at Edinburgh is internationally leading in for compiler and architecture co-design and optimisation – this will form the backbone to this project.
Candidate Profile
Suitable candidates will have a strong first degree in Computer Science or Mathemeatics and a strong interest in parallel programming, design space exploration, optimizing compilers or machine learning. The exact topic of the PhD is flexible depending on the candidate’s interests. We are looking for the brightest minds to pursue research in a cutting-edge arena. Candidate with Masters degrees or significant insdustrial experience are also strongly encouraged to apply.
Funding and further information
These studentships will cover all tuition fees and provide a tax-free stipend at the EPSRC rate. Students receive funding for a full EPSRC studentship for 4 years.
Contact Professor Michael O’Boyle (mob@inf.ed.ac.uk) to informally discuss the project further before applying.
Formal Specification and Proofs for TrustZone in ARMv8-M
Sponsored by ARM
The increasing complexity and connectivity in microcontroller devices has prompted the development of rotection mechanisms to improve reliability and security. ARM’s TrustZone for ARMv8-M provides a separate “secure world” execution mode to enable features such as secure firmware updates, safe integration of code from multiple suppliers and controlled access to privileged peripherals. It is designed in a different way to TrustZone on other ARM architectures, with greater hardware support for low-resource microcontrollers.
This project would study the low-level instruction set design of TrustZone for ARMv8-M, devising formal specifications describing the security properties that hold at the instruction level and proofs that these provide the intended protection against low-level attacks. The specifications would also aim to provide a basis for higher-level proofs about the security properties of software that runs in the secure world.
The exact details of the project are flexible depending on the candidate’s interests and background and ARM’s progress and priorities at the time of starting.
Funding and further information
This studentship will cover all tuition fees and provide a tax-free stipend at the EPSRC rate. Students receive funding for a full, 4-year EPSRC studentship, plus an additional 3.5k per annum and a funded internship.
Ccontact Dr Ian Stark (Ian.Stark@ed.ac.uk) or Professor David Aspinall (David.Aspinall@ed.ac.uk) to informally discuss the project further before applying.
Formal Verification of Security Properties of the mbed OS uVisor
Sponsored by ARM
The mbed OS uVisor is a core security component for ARM’s mbed IoT platform. It creates isolated security domains on Cortex M3, M4 and M7 microcontrollers with a Memory Protection Unit (MPU). The MPU provides a small number of memory regions of limited size. On top of these the uVisor provides a more flexible compartmentalisation using separate security domains (“Secure Boxes”), configured with ACLs.
This project will apply theorem proving methods to help define and then verify correctness and security properties of the uVisor implementation, building on previous work which has constructed instruction set models and decompilation techniques, and extending with formal modelling of exception handling and memory protection. Example verification goals include correctness aspects, such as ensuring that MPU regions get correctly reallocated on demand (ensuring progress/availability), or perhaps higher-level properties such as control-flow integrity (ensuring certain attacks are not possible in client code).
The exact details of the project are flexible depending on the candidate’s interests and background and ARM’s progress and priorities at the time of starting.
Funding and further information
This studentship will cover all tuition fees and provide a tax-free stipend at the EPSRC rate. Students receive funding for a full, 4-year EPSRC studentship, plus an additional 3.5k per annum and a funded internship.
Contact Dr Ian Stark (Ian.Stark@ed.ac.uk) or Professor David Aspinall (David.Aspinall@ed.ac.uk) to informally discuss the project further before applying.
GPU Compiler Optimisation
Sponsored by ARM
GPUs were initially designed as dedicated hardware accelerators for a graphics pipeline. Since then, they have expanded their usage to a much wider class of application. GPUs are widely recognised as having the potential to deliver power-efficient, high performance. However, achieving this potential is difficult, due to rapidly evolving architecture and the increasing diversity of applications.
This project will investigate compiler optimisation to improve the performance portability of GPUs. It will explore the impact compiler/runtime optimisation has on existing OpenCL and graphics shader workloads as well as emerging applications from Computer Vision. The area of research is broad and will depend on the student’s interests ranging from high-level parallelisation and mapping of work-flows to low-level instructions scheduling for GPU ISAs.
This project is sponsored by ARM, the world’s leading semiconductor IP company, where there is a unique opportunity to work under the hood on next generation GPUs.
The exact details of the project are flexible depending on the candidate’s interests and background.
Funding and further information
This studentship will cover all tuition fees and provide a tax-free stipend at the EPSRC rate. Students receive funding for a full, 4-year EPSRC studentship, plus an additional 3.5k per annum and a funded internship.
Contact Professor Michael O’Boyle (mob@inf.ed.ac.uk) to informally discuss the project further before applying.
Smart Parallelisation and Architecture Support
Sponsored by ARM
From data centres to smartphones, today’s computers are based on parallel multi-cores. They offer the potential for scaled performance while avoiding excessive power density. Efficient use of such systems, however, relies on both sufficient program parallelism and accurate exploitation of that parallelism by the hardware. Given that much software consists largely of sequential code and computer hardware is constantly changing, efficient parallelisation is a real challenge.
This project is concerned with parallelism discovery using static and dynamic analysis followed by portable smart mapping of parallelism to hardware. There are multiple areas for research which include:
– Parallel/vector trade-off in large systems
Here we are concerned with discovering parallelism and exploring what type of hardware best exploits it. For example, is using multiple-core or vector/SIMD best in terms of performance and energy? Integrating this into partial parallel MPI programs, providing support for speculation and dynamic allocation of parallelism are other areas of interest.
– Exploring hardware design to exploit available parallelism
Using dynamic analysis it can be shown that programs have considerably parallelism that cannot be exploited efficiently. Here, we are concerned with investigating how future hardware should be designed. By changing costs or redesigning micro-architecture, this work will investigate to what extent hardware can efficiently support the exploitation of more aggressive program parallelisation.
This project is sponsored by ARM, the world’s leading semiconductor IP company. where there is an opportunity to work across research groups.
The exact details of each project are flexible depending on the candidate’s interests and background.
Funding and further information
This studentship will cover all tuition fees and provide a tax-free stipend at the EPSRC rate. Students receive funding for a full, 4-year EPSRC studentship, plus an additional 3.5k per annum and a funded internship.
Contact Professor Michael O’Boyle (mob@inf.ed.ac.uk) or Dr Björn Franke (bfranke@inf.ed.ac.uk) to informally discuss the project further before applying.
Specialised Compiler/Architecture Design for Computer Graphics
Sponsored by ARM
Computer Graphics is the driving force behind much of the innovation in mobile devices. The need for high quality, low power realtime graphics has led to specialised Graphics Processing Units (GPUs). Such processors exploit massive parallelism and have specialised units for different parts of the graphics pipeline. Despite graphics being the driver behind GPUs, there is a large disconnect between the games developers, graphics frameworks, shader compilers, runtime systems and graphics hardware.
This project will investigate cross-stack optimisation of graphics processors. It will explore how existing hardware and system software optimisation fits emerging graphics applications. It will examine how to deliver next generation graphics in a real-time energy constrained setting. The area of research is broad and will depend on the student’s interests ranging from e.g. optimising ray-tracing to low-power hardware.
This project is sponsored by ARM, the world’s leading semiconductor IP company, and involves working with their media and graphics group.
The exact details of each project are flexible depending on the candidate’s interests and background.
Funding and further information
This studentship will cover all tuition fees and provide a tax-free stipend at the EPSRC rate. Students receive funding for a full, 4-year EPSRC studentship, plus an additional 3.5k per annum and a funded internship.
Contact Professor Michael O’Boyle (mob@inf.ed.ac.uk) to informally discuss the project further before applying.
Industry Sponsored Scholarships through the School of Informatics
CDT in Pervasive Parallelism industry partners Microsoft Research and Google also provide research scholarships to students in the School of Informatics. CDT PPar applicants and students are good candidates for these. Further information is available on the School of Informatics website: