Crash doing FSM

Questions
#1

Doing a simple Frozen String Calculation for a tutorial. When the calculation terminates, I see this at the command line:

exported envs:   QC QCAUX QCSCRATCH QCRUNNAME QCFILEPREF QCPROG QCPROG_S GUIFILE
/work/QChem6p3/bin/qchem: line 129: 3117307 Segmentation fault      (core dumped) ${QCPROG_S} ${inp} ${scr}
Error in Q-Chem run part 1
Error in the Q-Chem run

In the output file I see this:

Grid used for NLC:  SG-1 standard quadrature
 tid: 0x0 [2025.07.06-22:35:34.286] [INFO] Re-init GPU for small molecule mode
 tid: 0x0 [2025.07.06-22:35:34.289] [ERROR] Assertion failure: f
at /quantum/src/ocl_handler/device_info.cpp:115 in getDeviceID
/sys/bus/pci/devices/0000:e1:00.0/device does not exist

Here’s the input file:

$comment
DFT Freezing String Method for CH3Cl + OH- -> CH3OH + Cl-
$end

$molecule
-1 1
C        0.056711930     -0.062860300     -0.303673849
H        0.056711930      0.967579526     -0.655539121
H        0.949098876     -0.578080553     -0.655539147
H       -0.835675015     -0.578080553     -0.655539147
Cl       0.056711930     -0.062860186      1.476573846
O        0.070571530     -0.036498163     -3.303526004
H        0.070571530     -0.036498163     -4.280996026
****
C        0.780223323     -0.029319122      0.012029535
H        1.133876761     -0.554918783     -0.882131800
H        1.183123054      0.982801880      0.011023853
H        1.133925777     -0.553157501      0.907212685
Cl       3.752539494     -0.262893322      0.344865342
O       -0.644136836      0.114730913      0.011940281
H       -1.046405722     -0.769871693      0.012146220
$end

$rem
JOBTYPE        fsm           ! Freezing String Method
FSM_NGRAD      4            ! Number of perpendicular gradients per node
FSM_NNODE      15           ! Number of nodes along the string
FSM_MODE       2            ! LST interpolation (recommended)
FSM_OPT_MODE   2            ! Quasi-Newton optimization
METHOD         wB97M-V      ! mp2 method
BASIS          def2-SVP     ! Basis set
MOLDEN_FORMAT  TRUE         ! Output geometries in MolDen format
GUI            2
$end

Any idea what’s going on? Any suggestions on how to fix this?

FSM string distribution
#2

This may be a GPU-related issue, which I cannot diagnose. Using Q-Chem 6.3 on CPUs only, your input ran successfully through 150 freezing string iterations before timing out overnight (because I only gave it 4h on a few processors).

#3

Thanks John. Should I contact the BrianQC people directly or should I go through qchem support?

#4

Not sure, but I don’t know that Q-Chem has access to BrianQC source code so maybe the latter.