[Skip to Content]

ICRAR joins CUDA Research Centre Program

Dr Dodson, who is currently partly funded by an Academic Partnership with NVIDIA, led the proposal supported by various groups in ICRAR/UWA. The proposal stressed the importance of High Performance Computing, and the role that Graphic Processing Units (GPUs) such as those manufactured by NVIDIA, could play in addressing the challenges that Radio Astronomy faces in the SKA-epoch.

NVIDIA-black-600_400

Radio Astronomy is fast becoming a High Performance Computing exercise, and GPUs offer an affordable way to provide that performance. Whether the GPU is used in: the search for new mysterious astronomical objects (such as the strange `Lorimer Burst’ discovered in radio observations but not at all understood), for new forms of signals (such as Gravity Waves), or for the processing of the torrent of data generated by the next- generation `Square Kilometre Array’ telescope, the collaboration between ICRAR and NVIDIA could make it possible to tackle complex challenges, which would result in a myriad of spin-offs.

Under the two programs NVIDIA has donated both hardware and cash, which is advancing research into: Fast Transients for the ASKAP CRAFT project (1), Gravity Wave Searches (2), Simulations on cosmological (3) and galaxy scales (4), High Speed Correlation for Single Dishes and for Arrays (5), and High Performance Computing in general.

Relevant Papers

  • Dodson et al, 2009, “Commensual Real-time ASKAP Fast Transient Survey”, ASKAP Science Survey Proposal
  • Chung et al, 2010, “Application of graphics processing units to search pipelines for gravitational waves from coalescing binaries of compact objects”, Cl. Q. Grav., 27, 135009
  • Duffy et al, 2008, “Dark matter halo concentrations in the Wilkinson Microwave Anisotropy Probe year 5 cosmology”, MNRAS, 390, 64
  • Bekki & Tsujimoto, 2011 “Formation of the Galactic bulge from a two-component stellar disc: explaining cylindrical rotation and a vertical metallicity gradient”, 416, 60
  • Wayth,  Greenhill, Briggs, 2008, “A GPU-based Real-time Software Correlation System for the Murchison Widefield Array Prototype” PASP, 121, 882, 857