Ingesting data at 3,600 GBytes/s, storing 10,000s of 200TB files using a 64PFlop supercomputer (64PFlop may be on the low side) and trying to understand the evolution of galaxies; might sound a little ludicrous but it’s what Kevin does for a living.
In this Education across the Nation presentation Kevin will describe the scale of The Square Kilometre Array (SKA) Science Data Processing (SDP) problem and how an international team of Astronomy, High Performance Computing, and Software Engineering specialists are trying to have a design completed by the end of 2015.
Some of the problems facing the team are purely technical, but bringing 21 organisations together from a dozen countries to design something of this size, scale and complexity presents its own set of challenges when it comes to Software Engineering, Modelling, Communications and team dynamics.
Kevin will also talk about his other passion - theSkyNet POGS and how machine learning will help with the huge datasets.
Presentations will be held in each State – click on your State to register today.
Research Associate Professor Kevin Vinsen joined the International Centre for Radio Astronomy Research (ICRAR) in late 2009. ICRAR is a joint venture between Curtin University and The University of Western Australia, with funding from the State Government of Western Australia.
Kevin’s focus is for the issues caused by the huge data sets that modern radio and optical astronomy generates. As a member of ICRAR’s Data Intensive Research Group, Kevin is working with colleagues from around the world to design the science data pipe for the Square Kilometre Array radio telescope – which, when constructed, will be the biggest telescope on the planet.
Kevin’s main research interests are high speed data ingest, developing methods for the automated classification of galaxies using multi-wavelength data, machine learning algorithms and modelling complex systems.
When he’s not dealing with super computers Kevin works on on a citizen science project called the PS1 Optical Galaxy Survey (POGS), a part of theSkyNet initiative (www.theskynet.org). Using the collective processing power of home computers POGS is helping astronomers and astrophysicists to calculate the spectral energy distributions from optical infra-red and ultraviolet images to produce the first public catalog of its kind. This will require 10’s of millions of CPU hours to calculate and 100’s of TBytes of storage.
Kevin considers himself one of the luckiest geeks on the planet. He is paid to do what he loves - astronomy and computing with some of the biggest baddest computers on the planet. No wonder he is always smiling.