Runtime Scheduling Techniques to Manage Hardware Resource Sharing of Parallel and Concurrent Programs

GLOBAL RESEARCH INSTITUTES

  • Runtime Scheduling Techniques to Manage Hardware Resource Sharing of Parallel and Concurrent Programs

Runtime Scheduling Techniques to Manage Hardware Resource Sharing of Parallel and Concurrent Programs

Principal Supervisor: Dr. Hans Vandierendonck     

Second Supervisor:   Dr. Ivor Spence                                                            

+ Project Description

With a rising trend in core counts in multi-core processors, it will be increasingly common to share the cores between independent computations, e.g., originating from distinct applications. This poses issues in particular for parallel and concurrent applications as application performance is not easily composed. Many hardware resource sharing effects may annihilate the performance of individual applications, such as CPU core sharing (CPU oversubscription), CPU cache sharing and memory bandwidth sharing. Resource sharing affects both performance and fairness. This makes it hard, or even impossible, to tune the performance of parallel and concurrent applications as the operating conditions are hard to predict.

+ Objectives

The objectives of this project are:

  • To identify key resource sharing issues resulting from the co-location of parallel and concurrent applications on high core-count multi-core processors.
  • To investigate and develop practical runtime scheduling techniques to optimize hardware resource sharing for a range of performance and fairness metrics.

+ How to Apply

Applicants should apply electronically through the Queen’s online application portal at: https://dap.qub.ac.uk/portal/

+ Contact Details

Supervisor Name: Dr. Hans Vandierendonck
Address:

Queens University of Belfast
School of EEECS
Computer Science Building,
18 Malone Road

Email:

h.vandierendonck@qub.ac.uk

Tel:

+44 (0)28 9097 4654