- Topic: Butterfly Effects – Holistic Dynamic Energy Efficiency Perspectives for Computing Systems Design
- Closing Date: apply as soon as possible
- Eligibility: UK Students, EU Students
- Funding: Annual tax-free stipend with annual increment, fees fully funded and travel bursary provided
- Hours: Full Time
The Leverhulme Centre for Material Social Futures
Lancaster University’s Leverhulme Doctoral Training Centre in Material Social Futures is a major new strategic collaborative partnership between two of the university’s recently formed research Institutes – the Institute for Social Futures and the Material Science Institute. Based in the School of Computing and Communications (SCC) you will be part of a growing team of PhDs who will examine how to create more sustainable and socially beneficial futures, and who will be trained to engage in diverse aspects of materials discovery and the analysis of social and economic structures to achieve these ends. In short, the goal of PhDs in Material Social Futures will be to help produce futures that people want and the world needs.
Background to this PhD
We live in a world where energy demand continues to expand. Individuals and institutions do not just use more, they consume it in ever more diverse ways; and whilst there might be a willingness to be more sensitive to sustainable energy solutions, the increase in demand outstrips the pace with which new, more sustainable sources can be developed. This is particularly the case with computer systems and devices, where it is not just plugged and connected power that matters, but also for a portable battery provided sources. Smartphones succeed not just because of their smart apps, but because of the energy solutions they embody – in terms of the size and efficiency of the battery in relation to application demands. However, battery technologies use increasingly rare materials and substitutes do not always offer the same energy storage or capacity; and while computational power demands might be made more efficient, future smart devices might need to be quite radically reconceived. Just as this applies to smart devices, much larger computational systems such as Cloud computing might also need to be reconfigured to make energy provision and consumption key to their design.
ICT now consumes 10% of global electricity and is positioned to aggressively expand due to consumer demand and the rise of the Internet of Things (IoT) formed by 75 billion connected devices by 2025. Computing systems ranging from smartphones, laptops, to Cloud data centres, are composed by an assortment of software, hardware, battery, and cooling sub-systems.
Achieving high performance, energy-efficient computing is an active research area within computer science, with disciplines focusing on a singular sub-system of interest. Few approaches have addressed this holistically (all sub-systems together), nor considered the ‘butterfly effect’ that exists within sub-systems. For example, alteration to software patterns drives changes in resource usage, resulting in different energy consumption and heat rejection cooling patterns. The challenge of studying this butterfly effect is exhuberated when considering dynamic sub-system energy profiles from heterogeneous software patterns, architectures, battery technologies, and cooling types. If this is one perspective on energy, another has to do with how new, more ecological friendly batteries, have different energy production profiles and these too might affect how computer systems function efficiently.
The aim of this project is to study such things as the butterfly effect or the dynamics of power outputs with new battery technologies and their consequences for computing systems. If we are able to uncover the butterfly effect phenomena, would it be possible for computing systems to holistically coordinate and optimise their operation intelligently through the use of machine learning embedded within the Operating System in such a way as to optimise energy use? If battery output modulations are a key consequence of post-cobalt chemistries, what implications does this have for tetherless systems and their operating systems? Approaches to these and related problems, would not be limited to singular computing devices and systems and could encompass massive-scale distributed systems.Moreover, answers to these concerns will be deepened by team-based collaborations with chemists working on new battery technologies. The purpose of the research is to propose future computing architectures that would significantly improve the energy implication consequences of future computing architectures.
- Cover full payment of academic fees (at the standard RCUK rate)
- A Maintenance Stipend (£14,777 pa)
- Access to a Research Training Support Grant (RTSG) for reimbursement of research-related expenses including – but not limited to – conference attendance, training courses and equipment of at least £800 pa. Additional research costs (such as entailed in fieldwork) will be supported as appropriate
- Access to a range of training and development provided by the Material Social Futures PhD Programme, Sociology Department, the Faculty of Arts and Social Sciences, the Institute for Social Futures and Lancaster University
- The Material Social Futures PhD programme will offer internships (including international placements) in the second and or third year of training
Candidates should have a 1st class degree (or equivalent) in computer science or a closely related discipline with an interest in machine learning, energy-efficiency, operating systems, and distributed systems. Good programming skills are essential. Hands-on experience with computing architectures, battery technologies, and cooling systems are beneficial, but not required.
Please apply online via the University Postgraduate Admissions Portal with:
- A CV (2 pages maximum)
- Cover letter
- University grade transcripts
Note that no proposal is required as part of the application, though evidence of research vision and relevant background knowledge on the state of the art in this area is strongly encouraged. You should clearly state on your application that you are applying for a funded PhD opportunity on “Butterfly Effect – Holistic Energy-efficient Computing Systems ”.
We very much welcome informal queries about this opportunity, which should be directed to Dr Peter Garraghan (firstname.lastname@example.org)