-
Prof. Oliver Hahn (Lagrange/UCA/OCA)01/10/2019, 14:00
-
Mr Bournaud Frederic (CEA Saclay)01/10/2019, 14:20
-
Dr Jeremy Blaizot (CRAL)01/10/2019, 14:30
Community codes such as RAMSES face similar challenges regarding maintenance, upgrades, documentation, and user support, as classical observing instruments. They further face the issue of rapidly evolving platforms (e.g. computer centres buying GPUs), and have to manage a posteriori community developments. The French RAMSES community has been lobbying for RAMSES (and simulation codes in...
Go to contribution page -
Dr Neil Vaytet (European Spallation Source, Copenhagen, Denmark)01/10/2019, 15:20
I left the field of astrophysics a year ago and started working as a full-time software developer for the European Spallation Source, a large neutron physics facility current under construction in Lund (Sweden). I am involved in a sizeable project counting about 30 core developers, over 50 more casual developers, and ~2000 users. I will share my experiences on the methods we use to coordinate...
Go to contribution page -
Mr San Han (Yonsei University)01/10/2019, 15:40
Classic RAMSES code is purely written with MPI and not optimized for large simulations that use more than thousands of cores. We try to overcome this issue by adding OpenMP feature to RAMSES, implementing hybrid parallelism that takes advantage of both distributed and shared memory models. With the latest developed OpenMP version, we ran a small cosmological test simulation and it showed a...
Go to contribution page -
Prof. Romain Teyssier (University of Zurich)01/10/2019, 16:00
Choose timezone
Your profile timezone: