Hello AMBER Users,
I am experiencing an issue with minimization using sander.MPI. It gives a
segmentation fault error, and when I go to read the corresponding
minimization mdout file, the last lines are in Section 3:
*--------------------------------------------------------------------------------
3. ATOMIC COORDINATES AND
VELOCITIES--------------------------------------------------------------------------------default_name
begin
time read from input coords = 0.000 ps Number of triangulated 3-point
waters found: 1396247*
It may be worth mentioning that this simulation system (which I refer to as
*sdstest1* in all files) contains 125 surfactant molecules, 125 sodium
counterions, and 1.39 million TIP3P water molecules in a ~346 angstrom cube
(to create a 5 mM surfactant aggregation system)
I have also run a second test (*sdstest16*) corresponding to an 80 mM
surfactant solution (125 surfactant molecules, 125 sodium counterions, and
87,040
sdstest1.inpcrd
<
https://drive.google.com/file/d/1I1IJCP5xC1INgzefwOxwR1emm6x_PbIQ/view?usp=drive_web>
sdstest1_hmr.prmtop
<
https://drive.google.com/file/d/1O9-s6wXlpjnloCeXzTNMIp3Hbq4X1Dtd/view?usp=drive_web>
sdstest1_minimize.in
<
https://drive.google.com/file/d/1wusQN7tbqZyU1QfYAe9Nqs6S2JlFnE5I/view?usp=drive_web>
sdstest1_minimize.out
<
https://drive.google.com/file/d/1sp-MenZOKY4ncN4rwtwmGkM8LpY2ZAaZ/view?usp=drive_web>
sdstest16.inpcrd
<
https://drive.google.com/file/d/1kmrPu_gDbvbFJ55MZP6-7GgefligRBN3/view?usp=drive_web>
sdstest16_hmr.prmtop
<
https://drive.google.com/file/d/1QxHFcW4rRvHTkV36sdiZ2LKHPdYUunOw/view?usp=drive_web>
sdstest16_minimize.in
<
https://drive.google.com/file/d/1eSKHdrWrsdq3j6PBd0LxNSIs0UO1SIIu/view?usp=drive_web>
sdstest16_minimize.out
<
https://drive.google.com/file/d/1KQuSfo66hu-LrYs5Bz29gZlIykDlkfN6/view?usp=drive_web>
simulate_sdstest1_TEST_GPU0
<
https://drive.google.com/file/d/1_VT3bjwBfAqgIcgrxkOOgS_VXbfRhWpF/view?usp=drive_web>
simulate_sdstest16_TEST_GPU1
<
https://drive.google.com/file/d/1sKKyvh2yzO3oSZYR_DDHcn-M9r9MAerj/view?usp=drive_web>
TIP3P water molecules in a ~137 angstrom cube) , which has successfully
completed minimization in sander.
In the sdstest16 minimization mdout file (successful), its section 3
reads:
*--------------------------------------------------------------------------------
3. ATOMIC COORDINATES AND
VELOCITIES--------------------------------------------------------------------------------default_name
begin
time read from input coords = 0.000 ps Number of triangulated 3-point
waters found: 87040| Atom division among processors: | 0
11105 22208 33311 44414 55517 66620 77723| 88826 99929
111032 122135 133238 144341 155444 166547| 177650 188753 199856
210959 222062 233165 244268 255371| 266495 Sum of charges from
parm topology file = 0.00000043 Forcing neutrality...| Running
AMBER/MPI version on 24 nodes| WARNING: The number of processors is not
a power of 2| this may be inefficient on some systems.*
This leads me to believe that the segmentation fault for *sdstest1*
occurred while dividing atoms among processors. I have ruled out the
possibility that there are fewer available processors than requested (24),
as I ran sdstest1 on my computer which had 48 available processors at the
time of the run.
I have included all .inpcrd, .prmtop, minimization .mdin, minimization
.mdoutand command files for both sdstest1 and sdstest16. Any help/insight
as to why sdstest1 is failing minimization is greatly appreciated.
Thank you for your continued help and support!
Nathan D. Black
_______________________________________________
AMBER mailing list
AMBER.ambermd.org
http://lists.ambermd.org/mailman/listinfo/amber
Received on Thu Aug 04 2022 - 13:32:59 PDT
