Postgrad LIVE! Study Fairs

Birmingham | Edinburgh | Liverpool | Sheffield | Southampton | Bristol

Wellcome Trust Featured PhD Programmes
University College London Featured PhD Programmes
University of St Andrews Featured PhD Programmes
University of Kent Featured PhD Programmes
University of Manchester Featured PhD Programmes

Evolutionary optimization of streaming applications on many-core systems

This project is no longer listed in the FindAPhD
database and may not be available.

Click here to search the FindAPhD database
for PhD studentship opportunities
  • Full or part time
    Dr G Tempesti
  • Application Deadline
    No more applications being accepted
  • Funded PhD Project (European/UK Students Only)
    Funded PhD Project (European/UK Students Only)

Project Description

Imagine a many-core system with thousands or millions of processing nodes that gets better and better with time at executing an application, “gracefully” providing optimal power usage while maximizing performance levels and tolerating component failures. An ongoing EPSRC project (GRACEFUL, in collaboration with the universities of Manchester and Southampton) is developing a custom many-core platform, based on a network of ARM processors and Xilinx FPGAs, specifically design to support adaptive mechanisms for on-line optimization.

The GRACEFUL approach relies on the operation of a distributed set of monitoring nodes that constantly collect information about the application currently running in the system (within application nodes) and implements local control processes for fault tolerance, power minimization, and/or performance optimization. Dedicated hardware resources are provided within the nodes to support these processes, which can involve mechanisms such as task migration, task duplication, power management, DVFS (Dynamic Voltage and Frequency Scaling), etc.

The proposed PhD project aims at leveraging the dedicated hardware resources in the GRACEFUL platform and the search capabilities of bio-inspired techniques based on evolutionary algorithms to implement self-optimizing applications. A particularly interesting subset of applications in this context is that of pipeline or streaming applications, computationally-intensive applications where kernel functions are applied in sequence to the data stream and which exhibit both data parallelism and data locality, which include many common algorithms in domains such as audio and video processing, cryptography, etc. An appropriate benchmarking application, with possible impact on the state of the art in the domain, could be character recognition applied to handwritten and historical documents.

For further details on this opportunity and how to apply, please visit the Department of Electronics website at:

Funding Notes

The studentship will cover the tuition fee at the home/EU rate (£4,165 in 2016/17) and a stipend at the standard research council rate for a period of 3 years (£14,057 in 2015/16) starting in October 2016.

Applicants should have or be about to graduate with at least an upper second class honours degree in Electronic and Electrical Engineering, Physics, Computer Science, Mathematics or a related subject.

Applicants are invited to contact Dr Gianluca Tempesti ([Email Address Removed]) for more information.

FindAPhD. Copyright 2005-2018
All rights reserved.