Message: optical photons and CPU issues Not Logged In (login)
 Next-in-Thread Next-in-Thread
 Next-in-Forum Next-in-Forum

Question optical photons and CPU issues 

Forum: Processes Involving Optical Photons
Date: 09 Mar, 2012
From: Athreya Shankar <Athreya Shankar>

Hello all,

I am simulating light collection from my detector. But, I am encountering a strange problem. When the incident particles are protons (range 1-15 MeV), I am facing no problems and the simulations run fine. The same is true for low energy electrons ( of the order of 100s of keV). But once I go to the MeV range of electrons, the simulation runs for some number of events and then, I randomly get a segmentation fault.

I tried the same test case (100000 10 MeV electrons) on my PC (3 GB RAM) and on a cluster (8 GB RAM). I found that , in my PC, it gives a segmentation fault after around 1100-1500 events, whereas, on the cluster, it gets 'KILLED' after 45000-48000 events. So, I checked the memory usage during run-time on my PC. To my surprise, only around 500 MB of the 3 GB was being used up. So, clearly it is not a memory related issue. But it is not physics or geometry related either because of the large difference in the performance of my PC and the cluster. Further, in all these events, my aim is just to get the number of photons produced in every event, so I kill all the photon tracks using StackingAction.

Is there any other parameter, that I must set so that the program runs through the desired number of events?

Any help is greatly appreciated,

Regards, Athreya

 Add Message Add Message
to: "optical photons and CPU issues"

 Subscribe Subscribe

This site runs SLAC HyperNews version 1.11-slac-98, derived from the original HyperNews