[gmx-users] g_cluster crashes on large structures?

[Mikko Huhtala]

I was trying to run g_cluster on a 2 ns trajectory of a largish system
of 55095 atoms (protein dimer in generous water). I selected only
alpha carbons for superposition in the cluster analysis. I get the
following error after g_cluster has asked for the superposition and
output groups and starts to work:


Fatal error: calloc for d1[i] (nelem=55095, elsize=4, file g_cluster.c, line 1172): Cannot allocate memory


There is 512 Mb of RAM and 1 Gb of swap on the system (Linux
2.4.17). Is this a bug or should I be in the market for more memory?

Mikko