Integrated circuit design has been among the drivers of technology in the past few decades, from the ubiquitous processors in computers and smartphones to everyday home and industrial appliances. As advancement in the semiconductor technology industry has rapidly grown, the complexity of circuit design and optimization has exponentially increased, leading to longer development cycles at a significantly higher fabrication cost. This challenge is especially exacerbated in mixed-signal design where digital and analogue signals coexist, such as data-converters, transceivers, and image sensors. Unlike traditional optimisation methods, Machine Learning (ML) based algorithms have the potential to accelerate optimisation for complex circuits with a large number of transistors. An ML-based algorithm could enhance the circuit design process by performing most iterative tasks in an electronic design automation software, extracting design patterns using supervised learning, and apply reinforcement learning in areas where less circuit data is available.
This project explores transistor-level circuit design and optimisation based on machine-learning techniques for image sensors. In recent years, the use of image sensors has gone beyond photography in smartphone cameras. These sensors now have a wide variety of capability to measure various classical and quantum properties of light, including photon counting, polarisation, and travel time. These sensors are now deployed in a multiple industry sectors, from autonomous vehicles, personal healthcare to helicopters beyond our planet. Each image sensor consists of several components, including pixels, analogue-to-digital converter, phased-locked loops. Each of these design blocks requires transistor-level optimisation for performance and power efficiency.
This project is in partnership with Cadence Design System, the leader in electronic design and automation within the integrated circuit industry. The PhD student has the opportunity of top-up funding and industrial mentorship from Cadence R&D during their PhD. Additionally, further collaboration opportunities with local and international industry partners are available. The student will work in a vibrant community of engineers and scientists within The School of Engineering.
The University of Edinburgh has been at the forefront of integrated circuit design, especially in solid-state imaging technology, since the early 1980s gaining an international reputation for research in early CCDs and then in the first MOS analogue sensors. This culminated in the design and demonstration of the world's first single-chip CMOS video camera, a University Spin-out company, close links to local and international industry with impacts in medical devices, and detector technology is commercialising in time-of-flight distance sensors for mobile phones and DIY.
Please contact Dr Danial Chitnis (firstname.lastname@example.org) for pre-application enquires regarding the topic of the PhD studentship.
The University of Edinburgh is committed to equality of opportunity for all its staff and students, and promotes a culture of inclusivity. Please see details here: https://www.ed.ac.uk/equality-diversity
• Undergraduate degree in Physics, Engineering, or computer science
• Familiarity with programming languages including but not limited to Python, C, C++
• Self-motivated, able to work independently and collaboratively and meet deliverables.
• Familiarity with Electronic Engineering software
• Familiarity with Machine Learning tools and algorithms
• Familiarity with high performance computing
• Experience in data presentation and visualisation
• Experience of working in a multi-disciplinary team.
Minimum entry qualification - an Honours degree at 2:1 or above (or International equivalent) in a relevant science or engineering discipline, possibly supported by an MSc Degree. Further information on English language requirements for EU/Overseas applicants.
Tuition fees + stipend are available for Home students. Overseas applicants (Including EU applicants) are welcome to apply, but funding only covers the home rate.