Big Data with Space and Sound

Big Data with Space and Sound

The University of Oxford’s e-Research Centre is joining ASTRON, the Netherlands Institute for Radio Astronomy, and IBM Research in a collaboration to investigate high performing, but low-power exascale computing systems aimed at developing advanced technologies for handling the massive amount of data that will be produced by the large-scale projects, such as Square Kilometre Array (SKA).

High performing microservers will be researched for Computational Musicology and Computation in Natural Environments. The research is being conducted under the User Platform of DOME Project, DOME being an initial €32.9 million, five-year collaboration led by ASTRON and IBM Research with the financial support of the Dutch Ministry of Economic Affairs and from the Province of Drenthe.

The next generation of large scientific instruments, of which the SKA is a key example, requires a high-performance computing architecture and data transfer links with a capacity that far exceeds current state-of-the-art technology. To solve this unprecedented challenge, the DOME project is investigating and developing an emerging technologies roadmap for large-scale and efficient exascale computing, data transport and storage processes. The project has developed low-power microservers (processor chips and memory chips packaged together with cooling), which contain all of the essential functions of a modern server, in 1/10th of the size.

The SKA project is set to produce enough raw data that can be equated to 35,000 feature length movies per second, and the data rate in the second phase of the project will be ten times greater. Both phases of the SKA project have power constraints that are of a major concern.

It is a considerable challenge to manage, process and store such large datasets, within budgets and constraints, while ‘pushing the envelope’ of technology. System on a Chip (SoC) platforms are a growing trend for power efficient computation, using the concept of many small compute nodes, as progression from many-core processing. Creating a sustainable computational platform for processing data is one of the key challenges of the project.

Signal processing, from space to sound

Computational Musicology involves the provision of computational infrastructure to capture and process musical performances. When a musical performance is recorded, be it in the studio or on stage, a wealth of data is created. Low-powered exascale computing systems are being investigated to support the complex processes and workflows needed for the simultaneous capture of multiple audio tracks, automated feature extraction and music information processing, which are being combined with next-generation Linked Data and Semantic Web technologies to drive music information retrieval in a scholarly (musicological) or commercial context.

As a member of the DOME User Platform we plan to demonstrate how to embed the microservers in musical instruments and live performances, to capture and process information in real-time. This will enable musicians to capture the live publication of complex musical performance data and metadata, driving innovative practices in music publishing, and novel applications in human-human and human-machine collaboration and interactions. The university is expected to receive its first microserver in early 2016 with a prototype ready by year-end.

Another area of our interest is to create data centres that are embedded in their natural environment, making best use of their surroundings. Plants offer an opportunity to combine biology with computing, to provide self-maintaining, self-repairing and resilient infrastructures in arid landscapes. Our investigation of Computation in Natural Environments is building a foundation towards bio-based data centres that work with the biodiversity of their surroundings, providing cost-effective cooling, monitoring and infrastructure.

Introducing the SKA

The Square Kilometre Array (SKA) project is a large-scale global scientific endeavour, with around 100 organisations across 20 countries, coming together to build the world's largest radio telescope. The project is led by the SKA Organisation, a not-for-profit company with its headquarters at Jodrell Bank Observatory, near Manchester in the UK.

The SKA will see back to a time before the first star sparkled. Optical telescopes see the light from stars. Before stars existed there was only gas; a radio telescope with the sensitivity of the SKA can see back in time to the gas that existed before stars were even born.

The SKA will address a wide range of fundamental questions in physics, astrophysics, cosmology and astrobiology. It will be able to investigate previously unexplored parts of the distant Universe in unprecedented detail and map it hundreds of times faster than any current facility.   

The SKA will be built in Southern Africa and Australia. There will be 3,000 dish antennas, each about 15m in diameter, as well as two other types of radio antennas (low and mid-frequency). The antennas will be arranged in five spiral arms and the dishes will extend up to 3,000 km from the centre region. Construction of the SKA is expected to begin in 2017 and conclude in 2024.

SKA Telescope: http://www.skatelescope.org

Introducing Computational Musicology

A wealth of music and music-related information is now available digitally, offering tantalising possibilities for digital musicologies. These resources include large collections of audio and scores, bibliographic and biographic data, and live performance ephemera -- not to mention the 'hidden' existence of these in other digital content. With such large and wide-ranging opportunities come new challenges in methods, principally in adapting technological solutions to assist musicologists in identifying, studying, and disseminating scholarly insights from amongst this 'data deluge'.

Computational musicology supports and enhances traditional music scholarship, enabling musicological explorations at massive scales afforded by this abundance of digital content. Low-powered exascale computing systems and next-generation Linked Data and Semantic Web technologies will allow us to combine and publish related portions of this data in real-time, as well as to connect to information beyond the music, from data describing the settings of the mixing desk to information on audience reception and cultural impact.

About ASTRON

ASTRON is the Netherlands Institute for Radio Astronomy, part of the Netherlands Organisation for Scientific Research (NWO). Its mission is to make discoveries in radio astronomy happen, via the development of novel and innovative technologies, the operation of world-class radio astronomy facilities, and the pursuit of fundamental astronomical research.

ASTRON: http://www.astron.nl/

About IBM Research

Now in its 71st year, IBM Research continues to define the future of information technology with more than 3,000 researchers in 12 labs located across six continents.

Scientists from IBM Research have produced six Nobel Laureates, 10 U.S. National Medals of Technology, five U.S. National Medals of Science, six Turing Awards, 19 inductees in the National Academy of Sciences and 20 inductees into the National Inventors Hall of Fame. For more information, please visit http://www.research.ibm.com.

More information

Read this whitepaper on the DOME project:  http://www.scribd.com/doc/125147649/Ultimate-Big-Data-Challenge.

See also, Big Bang meets Big Data , Huffington Post, 11 Nov. 2013.