Hi, I wasn't sure if bugs were intended to be posted here or to the bug forum. I am reproducing here though: Hi, I believe there is a dangling pointer somewhere in the CP2K code that is very reproducible with my input file which I can post upon request. The unchecked version of the code seems to fail at random steps during the BO MD run. The check version of the code gives the following error: forrtl: severe (408): fort: (7): Attempt to use pointer FORCE when it is not associated with a target Our computer scientist tells us that the following statement is needed somewhere in the CP2K code: if (.not. associated(FORCE)) then print *,'Error, FORCE pointer not associated with a target)' return endif The version of the code that we are using is dated May 8, 2007. If someone believes that it is fixed in a recent CVS version, we can go ahead and test. -- Nichols A. Romero, Ph.D. 1613 Denise Dr. Apt. D Forest Hill, MD 21050 443-567-8328 (C) 410-306-0709 (O)