Hi Dr. David, good evening.
> You need to look at the Energy.dat.dif file, and perhaps run the test
> case
> by hand to learn more. You can ignore if you are not planning to use
> the
> xray facility in pmemd.
There is no difference because the test did not generate any output, so
it seems. I did try to execute ir directly, to no use:
##################
$ cd /usr/local/chem/amber23/test/cuda/xray/480d
$ ll
total 40
-rw-r--r-- 1 root root 0 nov 11 05:52 Energy.dat
-rw-r--r-- 1 root root 1475 nov 11 05:52 Energy.dat.dif
-rw-r--r-- 1 root root 1440 jan 28 2022 Energy.dat_DPFP.save
-rw-r--r-- 1 root root 1440 jan 28 2022 Energy.dat_SPFP.save
-rwxr-xr-x 1 root root 1153 jan 28 2022 Run.480d
-rw-r--r-- 1 root root 3498 nov 11 05:52 xray_md1.out
-rw-r--r-- 1 root root 16460 jan 28 2022 xray_md1.out.save
localhost:/usr/local/chem/amber23/test/cuda/xray/480d # ./Run.480d DPFP
yes
STOP PMEMD Terminated Abnormally!
diffing Energy.dat_DPFP.save with Energy.dat
possible FAILURE: check Energy.dat.dif
==============================================================
$ ll
total 40
-rw-r--r-- 1 root root 0 nov 14 00:46 Energy.dat
-rw-r--r-- 1 root root 1475 nov 14 00:46 Energy.dat.dif
-rw-r--r-- 1 root root 1440 jan 28 2022 Energy.dat_DPFP.save
-rw-r--r-- 1 root root 1440 jan 28 2022 Energy.dat_SPFP.save
-rwxr-xr-x 1 root root 1153 jan 28 2022 Run.480d
-rw-r--r-- 1 root root 3498 nov 14 00:46 xray_md1.out
-rw-r--r-- 1 root root 16460 jan 28 2022 xray_md1.out.sav
##################
Any suggestions on how to proceed from here?
> All of these failures are from mdgx.cuda -- do you have plans to use
> this
> program (it would be rather rare.)?
I would not get my hopes too high that I wouldn't at all use mdgx, but
however it might be true that the cuda version may not be needed in
comparison to the serial and parallel versions?
> It could be that mdgx.cuda has a problem with the most recent CUDA
> releases.
Could I ask Mr. Mingxuan Jiang, if he is still reading this thread, to
check it he finds the same issue or if tis just me It could be an
initial indicator of so.
> grep for "program error" in the log file. (Safest is to use grep -i to
> case
> case-insensitivity.)
I must not be getting enough sleep this days:
##############
grep -i error
/usr/local/chem/amber23/logs/test_at_cuda/2023-11-11_05-38-01.log
Backtrace for this error:
./Run.testCase.min: Program error
make[2]: *** [Makefile:101: test.pbsa.cuda] Error 1
make[2]: Target 'test.cuda.serial2' not remade because of errors.
1 tests experienced errors
##############
Any ideas on this one?
Thanks a lot!
Jones
---
Jones de Andrade
(jdandrade.iq.ufrgs.br)
DFQ/IQ/UFRGS
Lattes: http://lattes.cnpq.br/6675936210583999
Orcid: https://orcid.org/0000-0003-3429-8119
ResearcherID: https://publons.com/researcher/AAC-5337-2019/
_______________________________________________
AMBER mailing list
AMBER.ambermd.org
http://lists.ambermd.org/mailman/listinfo/amber
Received on Mon Nov 13 2023 - 20:00:02 PST
