Message boards :
ATLAS application :
What ATLAS simulations are we running?
Message board moderation
Author | Message |
---|---|
Send message Joined: 15 Nov 14 Posts: 602 Credit: 24,371,321 RAC: 0 |
Does anyone know whether we are doing ATLAS simulations for the already completed run, or for the new run after the upgrade is completed? I have not heard anything from the project managers on this. Thanks. |
Send message Joined: 13 May 14 Posts: 387 Credit: 15,314,184 RAC: 0 |
The current simulations are to help with the analysis of data from the run that was just completed (run 2). In the last 4 years, 45 Petabytes of raw data was recorded from the ATLAS detector and this has to be continuously processed and filtered for physicists to analyse. Due to the sheer volume of data and complexity of processing it, it will take literally years to analyse and extract all the relevant physics results, and simulated data is essential for these analyses. So we will continue to simulate data for run 2 for the foreseeable future and it will be probably at least one year before we start simulations for the next run (run 3) scheduled to start in 2021. For anyone interested in more details of simulation and why it is needed I recommend these three blog articles - they are from 2014 but still relevant today. |
Send message Joined: 15 Nov 14 Posts: 602 Credit: 24,371,321 RAC: 0 |
Thank you. That is just what I need. |
Send message Joined: 27 Sep 04 Posts: 6 Credit: 3,476,481 RAC: 0 |
Does Theory use the same concept for the simulations as ATLAS and CMS? I really like Theory for their statistics (I can easily keep track on the number of events being simulated...). ATLAS and CMS are sort of "blind" regarding your output, so you can't keep track beyond the RAC that BOINC gives you (which can be bogus since a WU could "finish" with the VM machine doing nothing). |
Send message Joined: 15 Jun 08 Posts: 2411 Credit: 226,413,747 RAC: 131,862 |
On one side ATLAS and CMS are very similar to Theory as they also produce various outputs at the console windows or in a browser if you press the "graphics" button of your BOINC client. On the other side, due to the nature of each app, those outputs are different and need interpretation. CMS for example runs subtasks inside the VM and always makes a 10 min idle break between the end of a subtask and the start of the next one. |
©2024 CERN