PMEMD.CUDA does not recognise plumed=1 flag

[Mauro Bringas]

Hi,

I read in the latest PLUMED documentation that pmemd.cuda in Amber18 already works with plumed, but (at least in my compilation) it does not recognise plumed=1 flag. Do I have to perform any additional task to make it work? 

thanks!

Mauro

[Giovanni Bussi]

Hi!

As of now you need to follow the plumed manual and patch amber before compiling it.

We are working on an integration similar to the one that you have for sander, but it has not been finalized yet

Giovanni

--
You received this message because you are subscribed to the Google Groups "PLUMED users" group.
To unsubscribe from this group and stop receiving emails from it, send an email to plumed-users...@googlegroups.com.
To view this discussion on the web visit https://groups.google.com/d/msgid/plumed-users/185acbc9-1f3e-49e6-8ee9-ee7bf7d3eec9%40googlegroups.com.

[Mauro Bringas]

Hi

i tried the "plumed patch -p" but it does not have any "amber" option... Is it available in plumed-2.5.2 version?

Thanks!

Mauro

El miércoles, 4 de diciembre de 2019, 13:52:43 (UTC-3), Giovanni Bussi escribió:

Hi!

As of now you need to follow the plumed manual and patch amber before compiling it.

We are working on an integration similar to the one that you have for sander, but it has not been finalized yet

Giovanni

On Wed, Dec 4, 2019 at 5:38 PM Mauro Bringas <maubr...@gmail.com> wrote:

Hi,

I read in the latest PLUMED documentation that pmemd.cuda in Amber18 already works with plumed, but (at least in my compilation) it does not recognise plumed=1 flag. Do I have to perform any additional task to make it work? 

thanks!

Mauro

--
You received this message because you are subscribed to the Google Groups "PLUMED users" group.

To unsubscribe from this group and stop receiving emails from it, send an email to plumed...@googlegroups.com.

[Giovanni Bussi]

I think you should download plumed from github

To unsubscribe from this group and stop receiving emails from it, send an email to plumed-users...@googlegroups.com.
To view this discussion on the web visit https://groups.google.com/d/msgid/plumed-users/9d4d9467-f2b2-402e-912c-1d15f4a10dac%40googlegroups.com.

[Mauro Bringas]

Thank you very much, that worked (at least pmemd.cuda recognizes the flags).

Now it fails, and writes these lines...

"pmemd.cuda: Plumed.h:2605: plumed_gcreate: Assertion `plumed_gmain.p==((void *)0)' failed.
PLUMED: Timestep: 0.002000
PLUMED: KbT: 2.494323
PLUMED: Relevant bibliography:
PLUMED:   [1] The PLUMED consortium, Nat. Methods 16, 670 (2019)
PLUMED:   [2] Tribello, Bonomi, Branduardi, Camilloni, and Bussi, Comput. Phys. Commun. 185, 604 (2014)
PLUMED: Please read and cite where appropriate!
PLUMED: Finished setup

Program received signal SIGABRT: Process abort signal.

Backtrace for this error:
#0  0x2B3FBAF44347
#1  0x2B3FBAF4495E
#2  0x3F4D63269F
#3  0x3F4D632625
#4  0x3F4D633E04
#5  0x3F4D62B74D
#6  0x3F4D62B80F
#7  0x708062 in plumed_gcreate
#8  0x708EEB in plumed_f_gcreate_static at Plumed.c:0
#9  0x708EB4 in plumed_f_gcreate_
#10  0x4B4715 in __runmd_mod_MOD_runmd
#11  0x4EF493 in MAIN__ at pmemd.F90:0"

Do you have any idea of what this means?

Thanks!

Mauro

El miércoles, 4 de diciembre de 2019, 15:14:15 (UTC-3), Giovanni Bussi escribió:

I think you should download plumed from github

On Wed, Dec 4, 2019 at 6:29 PM Mauro Bringas <maubr...@gmail.com> wrote:

Hi

i tried the "plumed patch -p" but it does not have any "amber" option... Is it available in plumed-2.5.2 version?

Thanks!

Mauro

El miércoles, 4 de diciembre de 2019, 13:52:43 (UTC-3), Giovanni Bussi escribió:

Hi!

As of now you need to follow the plumed manual and patch amber before compiling it.

We are working on an integration similar to the one that you have for sander, but it has not been finalized yet

Giovanni

On Wed, Dec 4, 2019 at 5:38 PM Mauro Bringas <maubr...@gmail.com> wrote:

Hi,

I read in the latest PLUMED documentation that pmemd.cuda in Amber18 already works with plumed, but (at least in my compilation) it does not recognise plumed=1 flag. Do I have to perform any additional task to make it work? 

thanks!

Mauro

--
You received this message because you are subscribed to the Google Groups "PLUMED users" group.
To unsubscribe from this group and stop receiving emails from it, send an email to plumed...@googlegroups.com.
To view this discussion on the web visit https://groups.google.com/d/msgid/plumed-users/185acbc9-1f3e-49e6-8ee9-ee7bf7d3eec9%40googlegroups.com.

--
You received this message because you are subscribed to the Google Groups "PLUMED users" group.
To unsubscribe from this group and stop receiving emails from it, send an email to plumed...@googlegroups.com.

To view this discussion on the web visit https://groups.google.com/d/msgid/plumed-users/9d4d9467-f2b2-402e-912c-1d15f4a10dac%40googlegroups.com.

[viktor drobot]

Hello! Could you please describe your setup: hardware, distro, kernel, compilers, CUDA version etc that you used for compilation of both AMBER and PLUMED? I've been using that patch for months on different software/hardware flawlessly. You could also try to compile AMBER with -fPIC flag and check if that worked

[Mauro Bringas]

Hi!

The configure options for amber already had the -fPIC flag...

This is the output of "uname -a"

Linux compute-0-0.local 2.6.32-504.16.2.el6.x86_64 #1 SMP Wed Apr 22 06:48:29 UTC 2015 x86_64 x86_64 x86_64 GNU/Linux

and the OS is Rocks (kernel version 2.6.32)

The graphic card is a GTX1080, amber is version 18 and PLUMED 2.6 (2.6 branch in git)

Thanks!

Mauro

[viktor drobot]

This is pretty old setup. Tell us please your compiler information. Could you try running your system on another software/hardware?

[Mauro Bringas]

Thank you Viktor! I tried the exact same procedure in a different (and more up to date) machine and it worked.

Best,

Mauro