Square Kilometre Array Project

Lead Research Organisation: University of Cambridge
Department Name: Physics

Abstract

This grant is in support of the detailed design work for the SKA to be undertaken at the University of Cambridge. The SKA has now entered the detailed design phase. The Office of the SKA Organisation published a request for proposals in 2013. Cambridge responded to this request as the lead of the Science Data Processor consortium to undertake the work package of the same name and a major part of the Low-frequency Aperture Array and Mid-frequency aperture array consortia work packages led by ASTRON. All responses to the RfP were evaluated and scrutinised by the SKAO and Board of the SKA Organisation. After negotiation with the SKA Organisation all of the consortia in which Cambridge lead or are involved were successful and awarded the work by the SKA Organisation. This application is for the funding to undertake this work as awarded by the SKA Organisation.

This submission is to support a further year of design work through to Critical Design Review submission in 2018.

Cambridge leads the Science Data Processor Consortium and associated work package. This work package will deliver the design for the complete software and computing processing system excluding the real-time control. As lead organisation Cambridge will manage the delivery across the entire consortium and critically provide the project manager, project engineer and project scientist. Key elements of the design that will be delivered directly from Cambridge include (but are not limited to) the overall architecture, system design, the prototyping of the software and hardware systems including the management of the open architecture lab, all management documentation, and the full documentation sets for preliminary and critical design reviews. In addition Cambridge manages the industrial contracts which form a critical element of the delivery.

Cambridge also has a major role in the Low-frequency Aperture array and Mid-frequency Aperture Array work packages and associated consortia. Here we provide the system engineer in both cases and have responsibility for the overall system design and design of the low-frequency aperture-array element itself, SKALA.

Planned Impact

The SKA is a major new global astronomy facility. It will be operated as an observatory with time awarded via competitive application for observing time. At the same time, the SKA project is a major technical challenge especially within the area of High-performance computing, "Big Data" technologies and energy efficient computing.

The output of the work to be supported by the application will be detailed design and prototyping work for the SKA through to Critical Design review. The delivery of the SKA itself which will follow from this design work will benefit the full academic astronomy community within the partner countries and more widely. In addition the key science goals for the instrument are at the core of modern cosmology and physics and the results will have significant impact beyond the astronomical community both within scientific community more widely but also to Society via our better understanding of the Universe and fundamental physics. We note in particular that SKA phase 1 will provide un-paralled views of the Epoch of deionisation when the first objects form and studies of pulsars within our Galaxy will enable theories of gravity to be tested in extreme conditions and detection of the expected stochastic background of gravitational waves. The UK leadership in key design areas will ensure the UK astronomical community is ideally placed to maximise return from the SKA when operational.

The design work will also have a much wider impact. The SKA is widely recognised as the next BigData challenge within science. As such the required design is pushing the boundaries of those technologies needed to process, analyse and store streaming data. These technologies are critical to the SKA, but are also of wide and general applicability across a broad range of market sectors from standard Data Centres to the analysis and management of data from sensor networks such as those associated with Smart Cities. Via collaboration with major industry partners in the ICT sector and the targeted contracts to be managed as part of this work, the design work to be undertaken will have much wider applicability across the IT market sector. In particular the SKA acts as a single point of focus for many companies both in terms of their own development work, but also as a project for wider collaboration within the sector. In these ways the ICT sector will be a direct beneficiary of the SKA design work. Areas where IP is likely to be developed as part of the design work include new software for file and object stress, new software management layers for Data Centres/HPC facilities especially for BigData challenges, new tools for data management, new analytical approaches for BigData. For these beneficiaries we foresee that the impact may occur in a a number of ways:
1. IP will be developed via this work either jointly or within the industry partners. The IP arrangements in place enable industry partners to exploit this IP while retaining necessary rites for the delivery of the SKA. If the IP is jointly developed appropriate sharing of the results of the impact will be put in place following standard procedures established via Cambridge Enterprise.
2. The SKA design work will provide a focus for some industry partners to develop modified or new products and services which will have wider impact.
3. Direct funding of design work will be undertaken as part of this work. Exploitation of IP derived in this way will be determined by the contractual arrangements with those companies winning competitive tenders while ensuring the delivery of the SKA design according to the SKA IP policy.

Further IP generated wholly within Cambridge will be exploited appropriately via interaction with the University's wholly owned Cambridge Enterprise subsidiary.
 
Description This work has delivered designs for the SKA in two key areas:

Science Data Processor: Architecture, software, software models and planning
Low-frequency Aperture Array: Antenna design, system design, modelling and planning
Exploitation Route Elements of the software are being carried forward by commercial companies

The work will enable companies in the UK to bid for SKA construction contracts in 2021 onwards
Sectors Aerospace, Defence and Marine,Digital/Communication/Information Technologies (including Software),Electronics,Other

URL http://www.skatelecope.org
 
Description This grant supports detailed design for the Square Kilometre Array (SKA) Radio Telescope. The aspects of the design undertaken in Cambridge are for the Science Data Processor and Low Frequency Aperture Array elements. The Science Data Processor design encompasses all of the software and computing hardware to deliver the data analysis for the SKA through to the end users. The SKA is widely regarded as an iconic Big Data Project for the next decade. The data analysis is not only complex but also of very large volume - near real time analysis of data arriving at over 14 TBytes per second is required. Final data volumes reach of order 300 PBytes per year which must be delivered to regional Science Centres for further analysis. The delivery from this work is a detailed analysis and design of this system which will meet these challenges. For the Low-Frequency aperture Array the main design of is a state-of-the-art antenna element, associated electronics and overall system design for the element. The main output from this research, when completed in 2017, will be a detailed design documentation. This is essential input to the next stage of the SKA project which will be the construction and implementation phase. A major intermediate milestone deliverable has been the documentation set for the rebaselining process including a preliminary architecture, performance analysis and cost estimates. Given the nature of the Big Data challenge for the SKA there is considerable overlap with other areas and the approaches to the analysis of large data volumes will have significant broad impact and applicability. Low energy compute is a major target for this work where we consider system-level optimisation for low energy computing. The output of the design for the Science Data Processor is an architecture for a novel high-data compute intensive system which will have general applicability/
First Year Of Impact 2018
Sector Digital/Communication/Information Technologies (including Software),Electronics,Other
Impact Types Societal,Economic

 
Description SKA Mid Frequency Aperture Array Consortium 
Organisation Aeon Astron B.V.
Country Netherlands 
Sector Private 
PI Contribution We are developing the design for a novel antenna to be considered for the SKA mid-frequency development.
Collaborator Contribution This is a consortium to develop the mid frequency aperture array for the SKA,
Impact Design documentation for a preliminary design review has been submitted to the SKA Organisation.
Start Year 2013
 
Description SKA Mid Frequency Aperture Array Consortium 
Organisation University of Manchester
Department School of Medicine Manchester
Country United Kingdom 
Sector Academic/University 
PI Contribution We are developing the design for a novel antenna to be considered for the SKA mid-frequency development.
Collaborator Contribution This is a consortium to develop the mid frequency aperture array for the SKA,
Impact Design documentation for a preliminary design review has been submitted to the SKA Organisation.
Start Year 2013
 
Description SKA Mid Frequency Aperture Array Consortium 
Organisation University of Oxford
Department Wellcome Trust Centre for Human Genetics
Country United Kingdom 
Sector Charity/Non Profit 
PI Contribution We are developing the design for a novel antenna to be considered for the SKA mid-frequency development.
Collaborator Contribution This is a consortium to develop the mid frequency aperture array for the SKA,
Impact Design documentation for a preliminary design review has been submitted to the SKA Organisation.
Start Year 2013
 
Description SKA SDP Consortium 
Organisation SKA Square Kilometre Array
Department SKA Science Data Processing Consortium
Country United Kingdom 
Sector Charity/Non Profit 
PI Contribution Cambridge led this consortium with Alexander as Lead, Coles as Project Manager, Nikolic as project Engineer, Ashdown as project scientist and Wortmann as project architect. The consortium delivered the design of the SKA Science Data Processor through to Critical Design Review for the SKA. The SDP is a new high performance data analytics system capable of scaling to at least 250 PFlop performance
Collaborator Contribution Cambridge led this consortium with Alexander as Lead, Coles as Project Manager, Nikolic as project Engineer, Ashdown as project scientist and Wortmann as project architect.
Impact The consortium delivered the design of the SKA Science Data Processor through to Critical Design Review for the SKA. The SDP is a new high performance data analytics system capable of scaling to at least 250 PFlop performance. This is multi disciplinary including physicists, astronomers, computer scientists and software engineers as well as project management and system engineering expertise
Start Year 2013
 
Title SKA SDP Algorithm Reference Library 
Description The software is a reference implementation of the imaging and calibration algorithms required for SKA imaging. It is written in python and will be delivered to the SKA Organisation as part of the deliverables form the design work. The latest version of the library has also significant performance and scalability itself in conjunction with the DASK execution framework / environment 
Type Of Technology Software 
Year Produced 2019 
Open Source License? Yes  
Impact This will be the first end-to-end reference library for radio astronomy software. It enables us to engage with industry in considering the next construction phase of the programme providing a reference for critical software elements. 
URL http://ska-sdp.org
 
Title SKA Science Data Processor Architecture 
Description This is a software architecture and associated prototype code for the SKA Science Data Processor. The software implements a novel architecture for a highly scalable data analytics engine able to scale to at least 250 PFlop performance. The design is highly innovative and supports extensibility and maintainability as key drivers as well as performance. 
Type Of Technology Software 
Year Produced 2019 
Open Source License? Yes  
Impact Although this product has been developed for the SKA it has wide potential applicability to any problem requiring at-scale data analytics 
URL http://ska-sdp.org
 
Title SKALA4 
Description The SKALA is a novel antenna design for low-frequency radio astronomy offering excellent performance over a wide bandwidth from 40-800MHz. The antenna is optimised for low-noise performance and sky coverage. The latest version of the antenna has significant design improvements including a flatter bandpass and better design for mass manufacture 
Type Of Technology Systems, Materials & Instrumental Engineering 
Year Produced 2018 
Impact This antenna design is currently the favoured design of antenna for the SKA low-freqency aperture arrays. In the final SKA approximately 3,000,000 antennas of this sort will be deployed. The antenna will also likely be adopted (or variants of it) for a 
 
Title Scientific OpenStack 
Description The software has been developed as part of the SKA Science Data Processor Platform. It adds new functionality to the OpenStack platform to enable high performance workflows and other monitoring. The software has been added back into the main OpenStack repository for general use. 
Type Of Technology Software 
Year Produced 2018 
Open Source License? Yes  
Impact This software has wide applicability outside of radio astronomy and the SKA project for which it is developed and at least one company, StackHPC are taking this forward in a commercial context 
URL http://ska-sdp.org