| |
 |
| Paper: |
GPU Implementation in Radio Astronomy: a “Giant Leap” into the SKA Era |
| Volume: |
541, ADASS XXXIII |
| Page: |
245 |
| Authors: |
E. De Rubeis; C. Gheller; G. Taffoni; L. Tornatore; G. Lacopo; D. Goz |
| DOI: |
10.26624/VMOV8865 |
| Abstract: |
With the incoming new generation of radio interferometers, the necessity of speed-up in computation is becoming tangible. For radio astronomy, the current
state of the art software is not capable of facing such huge data challenge, that requires the exploitation of HPC infrastructures and, in particular, accelerators. The code
presented in this paper has been developed to target this issue, aiming at fully run on
GPUs the most computationally-expensive steps of radio astronomy imaging. After a
brief introduction to the code workflow, we focus on a more accurate description of the
methodology and libraries used for the GPU off-loading. |
|
|
|
|
|
