Amazon Web Services (AWS) Research Grant 1

Breakthroughs in telescope, detector, and computer technology allow astronomical instruments to produce terabytes of images and catalogs; astronomy is facing a data explosion. The data sets produced cover the sky in multiple band widths, from gamma and X Ray, optical, infrared through to radio. With such vast quantities of data being archived, it is becoming easier to `dial up’ a piece of the sky, rather than waiting for expensive, scarce telescope time. This Amazon research proposal focuses on using astronomical data archives to calculate Photometric Redshifts on the Amazon Web Services stack.

• Funding: Amazon Researcher Grant Programme.
• Adam Barker (PI).
• Status: Ongoing.

Amazon Web Services (AWS) Research Grant 2

Gary McGilvary’s PhD proposes a framework that determines the optimal cloud configuration parameters for a given computation job. These parameters include: the number of instances, their type and size as well as the number of cores and processes per instance and the minimum allocations of memory, storage and network performance for applications. This research proposal allows Gary to run his framework over Amazon’s Elastic Compute Cloud (EC2).

• Funding: Amazon Researcher Grant Programme.
• Gary McGilvary (PI).
• Status: Ongoing.

Microsoft Windows Azure Academic Collaboration

We are exploring autonomic management techniques on the Microsoft Windows Azure platform.

• Funding: Gift via Microsoft.
• Adam Barker (PI).
• Status: Ongoing.

Completed Research Projects

Elastic Virtual Infrastructure for Research Applications (ELVIRA)

The ELVIRA project aims to enable researchers to simply and rapidly deploy, execute and monitor scientific software on elastic cloud computing infrastructures. Our focus is on the ‘long tail’ of scientific applications that do not currently benefit from the development of e Infrastructures for research but that are of immense scientific value as they are used by large communities of users. Their users often struggle to run them using traditional HTC or HPC compute resources for a variety of reasons including: the workload characteristics, a lack of technical skills or that they lack additional requirements such as collaboration support through shared desktop environments.

• Funding: EPSRC and JISC.
• Ian Sommerville (PI), Adam Barker and Alexander Voss (CI).
• Status: Complete.