Accelerated 3D Visualization of Mock Galaxy Catalogues for the Dark Energy Survey

We present a novel implementation of the visualization code Splotch(1), exploiting optimized atomic operations of modem GPU hardware found in todays heterogeneous supercomputers, and show its application to the visualization of mock galaxy catalogues for the Dark Energy Survey (DES)(2). These mock catalogues are generated by numerical simulation codes producing results which can then be directly compared with observations or validated against other models. As a use case we compare a catalogue based on the L-PICOLA(3) code with the MICE4 catalogue for DES. We present and discuss the potentiality of our optimized visualization algorithm as a debugging tool for L-PICOLA, in particular focusing on gaining rapidly an insight into potential anomalies on large scales and aiding in planning of appropriate remediation measures.