I Multicore World, 27-28 March 2012, Wellington, New Zealand
Abstract - The Square Kilometre Array (SKA) will undisputedly produce Big Data: one to 10 Petabytes/s, and will require Big Computing: ten's to hundred's of Petaflops. Unlike the LHC, this computing cannot occur in the Cloud since all data are needed to make an image. Hence SKA will require a monolithic supercomputer, perhaps even located out in the desert near the telescope. These performance numbers are about a factor of a ten thousand bigger than the current largest load - that from the Australian Square Kilometre Array Pathfinder (ASKAP). In 5 years, we have scaled the ASKAP imaging code over a factor of one thousand. This required 5 changes in algorithm. Each change introduced more complexity in order to fit within the constraints of flops, memory, and memory access. Scaling to SKA levels is therefore going to be a very substantial challenge not least because at some point, the accumulating algorithm and code complexity must be reduced. This must occur even against a background of changing computing architectures. I will discuss the implications for Legacy Code - Mutate or Die!
Dr. Tim Cornwell - Senior Software Scientist, Astronomy and Space Science, Project Lead for ASKAP. CSIRO, Australia
Tim Cornwell has worked at the forefront of radio astronomical data processing for over 35 years. He has a PhD (1980) from the University of Manchester in England where he worked on image processing algorithms for radio synthesis telescopes. His first significant contribution was the development of the self-calibration algorithm widely used in radio astronomy. In 1980, he moved to Socorro, New Mexico to work on the newly completed Very Large Array telescope run by the National Radio Astronomy Observatory. Over the 25 years at the NRAO, he made many contributions to radio astronomical techniques, including the key algorithms needed for wide fields of view: mosaicing wide fields, and wide projection for correcting a Fresnel diffraction effect for wide fields. In 1986, he used the CRAY-YMP at Digital Productions to produce the first 16 MPixel radio interferometric image. He also contributed in the areas of telescope design (for the Atacama Large Millimeter Array), telescope commissioning (the Very Long Baseline Array), observatory management, and software development. In 2004, he joined the Square Kilometre Array International Engineering Working Group, primarily to contribute towards computing and algorithms. In 2005, he moved to Australia to take the lead role in computing for the Australia SKA Pathfinder (ASKAP). Since then he has been heavily involved in all aspects of the development of ASKAP computing, and most particularly in the provision of high performance computing for the telescope. He is currently working on the development of algorithms for ASKAP in C++. He also has experience in Forth, FORTRAN, and C. He contributed substantially to the open source CASA and casacore packages. He is now starting to work on planning software and computing for the Square Kilometre Array.
Loading more stuff…
Hmm…it looks like things are taking a while to load. Try again?