Re: [gmx-users] parallel processing

[abhisek Mondal]

Hello Mark,

I had gone through the page. Not finding any solution there.

My architecture is:
Architecture:  x86_64
CPU op-mode(s):32-bit, 64-bit
Byte Order:Little Endian
CPU(s):16
On-line CPU(s) list:   0-15
Thread(s) per core:1
Core(s) per socket:8
Socket(s): 2
NUMA node(s):  2
Vendor ID: GenuineIntel
CPU family:6
Model: 45
Stepping:  7
CPU MHz:   2593.616
BogoMIPS:  5186.82
Virtualization:VT-x
L1d cache: 32K
L1i cache: 32K
L2 cache:  256K
L3 cache:  20480K
NUMA node0 CPU(s): 0-7
NUMA node1 CPU(s): 8-15


When i put "mpirun -np 64 gmx_mpi mdrun  -v -deffnm npt" it just runs on a
single node using 64 threads (16 availavle for a single node normally).

How am I supposed to do node distribution efficiently ? I'm trying to run
the job using 4 nodes and 16 threads.
A little suggestion would be highly appreciated.



On Fri, Dec 2, 2016 at 4:55 PM, Mark Abraham 
wrote:

> Hi,
>
> Please check out
> http://manual.gromacs.org/documentation/2016.1/user-
> guide/mdrun-performance.html.
> Your first sample command is woefully inefficient, but for em this likely
> doesn't matter.
>
> Mark
>
> On Fri, 2 Dec 2016 09:01 abhisek Mondal  wrote:
>
> > But if I want to run the same job in 4 nodes (available cores= 4*16) then
> > how would this work ?
> >
> > On Fri, Dec 2, 2016 at 2:20 PM,  wrote:
> >
> > > Hi Abhisek,
> > >
> > > You would need to use another version of gromacs with mpi rather than
> > > thread-mpi (installed by adding DGMX_MPI=ON to the cmake command). You
> > > could then use the following command
> > >
> > > mpirun -np 4 gmx_mpi mdrun -ntomp 16 -npme 0 -v -deffnm em
> > >
> > > I'm not sure why you are specifying -npme 0 but would suggest you don't
> > do
> > > this and instead let the number of separate pme ranks be set
> > > automatically.
> > >
> > > Best wishes
> > > James
> > >
> > > > Hi,
> > > >
> > > > I'm running gromacs on a cluster configuration as follows:
> > > > 1 node = 16 cores
> > > >
> > > > I'm able to use single node with "gmx mdrun -ntmpi 4 -ntomp 16 -npme
> 0
> > -v
> > > > -deffnm em" command.
> > > >
> > > > How can I be able to run on multiple node (I have 20 nodes
> available) ?
> > > > "-nt" is not doing good here.
> > > >
> > > >
> > > >
> > > > --
> > > > Abhisek Mondal
> > > >
> > > > *Research Fellow*
> > > >
> > > > *Structural Biology and Bioinformatics Division*
> > > > *CSIR-Indian Institute of Chemical Biology*
> > > >
> > > > *Kolkata 700032*
> > > >
> > > > *INDIA*
> > > > --
> > > > Gromacs Users mailing list
> > > >
> > > > * Please search the archive at
> > > > http://www.gromacs.org/Support/Mailing_Lists/GMX-Users_List before
> > > > posting!
> > > >
> > > > * Can't post? Read http://www.gromacs.org/Support/Mailing_Lists
> > > >
> > > > * For (un)subscribe requests visit
> > > > https://maillist.sys.kth.se/mailman/listinfo/gromacs.org_gmx-users
> or
> > > send
> > > > a mail to gmx-users-requ...@gromacs.org.
> > > >
> > >
> > >
> > > --
> > > Gromacs Users mailing list
> > >
> > > * Please search the archive at http://www.gromacs.org/
> > > Support/Mailing_Lists/GMX-Users_List before posting!
> > >
> > > * Can't post? Read http://www.gromacs.org/Support/Mailing_Lists
> > >
> > > * For (un)subscribe requests visit
> > > https://maillist.sys.kth.se/mailman/listinfo/gromacs.org_gmx-users or
> > > send a mail to gmx-users-requ...@gromacs.org.
> > >
> >
> >
> >
> > --
> > Abhisek Mondal
> >
> > *Research Fellow*
> >
> > *Structural Biology and Bioinformatics Division*
> > *CSIR-Indian Institute of Chemical Biology*
> >
> > *Kolkata 700032*
> >
> > *INDIA*
> > --
> > Gromacs Users mailing list
> >
> > * Please search the archive at
> > http://www.gromacs.org/Support/Mailing_Lists/GMX-Users_List before
> > posting!
> >
> > * Can't post? Read http://www.gromacs.org/Support/Mailing_Lists
> >
> > * For (un)subscribe requests visit
> > https://maillist.sys.kth.se/mailman/listinfo/gromacs.org_gmx-users or
> > send a mail to gmx-users-requ...@gromacs.org.
> >
> --
> Gromacs Users mailing list
>
> * Please search the archive at http://www.gromacs.org/
> Support/Mailing_Lists/GMX-Users_List before posting!
>
> * Can't post? Read http://www.gromacs.org/Support/Mailing_Lists
>
> * For (un)subscribe requests visit
> https://maillist.sys.kth.se/mailman/listinfo/gromacs.org_gmx-users or
> send a mail to gmx-users-requ...@gromacs.org.
>



-- 
Abhisek Mondal

*Research Fellow*

*Structural Biology and Bioinformatics Division*
*CSIR-Indian Institute of Chemical Biology*

*Kolkata 700032*

*INDIA*
-- 
Gromacs Users mailing list

* Please search the archive at 
http://www.gromacs.org/Support/Mailing_Lists/GMX-Users_List before posting!

* Can't post? Read 

Re: [gmx-users] PMF decompostion

[David van der Spoel]

On 02/12/16 23:40, Christopher Neale wrote:

PMF gives free energy, which can not be decomposed into VDW and Q without 
leaving a remainder, though I guess that doesn't affect your question. Enthalpy 
is simply the average potential energy plus the pressure volume component. So 
you can compute average potential energy VDW and Q components at each umbrella 
if you want (from the .edr files with gmx energy). Be aware that the 
statistical uncertainty on the enthalpy tends to be greater than the relevant 
differences on achievable simulation timescales (i.e., convergence is hard), so 
you'll want to evaluate error explicitly.

Agree with Chris here, that enthalpy and entropy are the 
thermodynamically meaningful components here. VDW and Q does not mean 
anything outside the force field definition since they can not be 
measured in any way. For more info on decomposing into enthalpy and 
entropy, see this paper: http://pubs.acs.org/doi/abs/10.1021/ct400404q




From: gromacs.org_gmx-users-boun...@maillist.sys.kth.se 
 on behalf of Salman Zarrini 

Sent: 02 December 2016 17:16:15
To: gmx-us...@gromacs.org
Subject: [gmx-users] PMF decompostion

Dear gromacs user,

I was wondering if something has been already implemented in gromacs to
make possible the potential of mean force (PMF) decomposition into the van
der waals and electrostatic energies?

Would you please also let me know if the desolvation energy originated from
protein-water or in my case solid surface-water interaction, will be taken
into account automatically when one calculates the PMF in aqueous solution
in gromacs?

Thanks.

Cheers,
Salman Zarrini
--
Gromacs Users mailing list

* Please search the archive at 
http://www.gromacs.org/Support/Mailing_Lists/GMX-Users_List before posting!

* Can't post? Read http://www.gromacs.org/Support/Mailing_Lists

* For (un)subscribe requests visit
https://maillist.sys.kth.se/mailman/listinfo/gromacs.org_gmx-users or send a 
mail to gmx-users-requ...@gromacs.org.




--
David van der Spoel, Ph.D., Professor of Biology
Dept. of Cell & Molec. Biol., Uppsala University.
Box 596, 75124 Uppsala, Sweden. Phone:  +46184714205.
sp...@xray.bmc.uu.sehttp://folding.bmc.uu.se
--
Gromacs Users mailing list

* Please search the archive at 
http://www.gromacs.org/Support/Mailing_Lists/GMX-Users_List before posting!

* Can't post? Read http://www.gromacs.org/Support/Mailing_Lists

* For (un)subscribe requests visit
https://maillist.sys.kth.se/mailman/listinfo/gromacs.org_gmx-users or send a 
mail to gmx-users-requ...@gromacs.org.

Re: [gmx-users] COM group of Membrane and Protein simulation

[Mijiddorj Batsaikhan]

Dear Justin,

Thank you very much.
I attached my mdp file. Peptide locates on membrane surface. I want to know
about interaction between membrane and peptide, membrane permeability of
peptide, amino acids contribution for the penetration, specially energetic
values between the membrane and the peptide.
Please suggest me with helpful options?
How can I chose COM groups?

I also copied my mdp file.


integrator  = md
dt  = 0.002
nsteps  = 2500
nstlog  = 5000
nstxout = 5000
nstvout = 5000
nstfout = 5000
nstcalcenergy   = 100
nstenergy   = 100
nstxout-compressed  = 100
nstxtcout   = 100
compressed-x-grps   = System
energygrps  = PROT MEMB SOL
;
cutoff-scheme   = Verlet
nstlist = 10
rlist   = 1.2
coulombtype = pme
rcoulomb= 1.2
vdwtype = Cut-off
vdw-modifier= Force-switch
rvdw_switch = 1.0
rvdw= 1.2
;
tcoupl  = Nose-Hoover
tc_grps = PROT   MEMB   SOL_ION
tau_t   = 1.01.01.0
ref_t   = 313 313 313
;
pcoupl  = Parrinello-Rahman
pcoupltype  = semiisotropic
tau_p   = 5.0
compressibility = 4.5e-5  4.5e-5
ref_p   = 1.0 1.0
;
constraints = h-bonds
constraint_algorithm= LINCS
continuation= yes
;
nstcomm = 100
comm_mode   = linear
comm_grps   = PROT   MEMB   SOL_ION ; which one is better to
the simulation?
;comm_grps   = PROT_MEMB   SOL_ION ; which one is better to the
simulation?
;
refcoord_scaling= com

--

>
> Message: 5
> Date: Fri, 2 Dec 2016 07:50:52 -0500
> From: Justin Lemkul 
> To: gmx-us...@gromacs.org
> Subject: Re: [gmx-users] COM group of Membrane and Protein simulation
> Message-ID: 
> Content-Type: text/plain; charset=windows-1252; format=flowed
>
>
>
> On 12/1/16 9:17 PM, Mijiddorj Batsaikhan wrote:
> > Dear gmx_users,
> >
> > I started simulation that a peptide on membrane. My peptide locates on
> the
> > membrane surface. I have two questions relating to the simulation.
> > (1)
> > When I start the simulation, I chose COM groups separately. Is this
> choice
> > okay? or May I need to chose COM group inseparately?
> >
>
> Your description is ambiguous; please post the actual .mdp contents.
>
> > (2)
> > During the simulation peptide is moving the edge of membrane. How can I
> > shift the peptide to the central section of the membrane? Can I use
> > -nojump, -center options of trjconv tool?
> >
>
> Yes - try it and see.  Also relevant is -fit transxy.
>
> -Justin
>
> --
> ==
>
> Justin A. Lemkul, Ph.D.
> Ruth L. Kirschstein NRSA Postdoctoral Fellow
>
> Department of Pharmaceutical Sciences
> School of Pharmacy
> Health Sciences Facility II, Room 629
> University of Maryland, Baltimore
> 20 Penn St.
> Baltimore, MD 21201
>
> jalem...@outerbanks.umaryland.edu | (410) 706-7441
> http://mackerell.umaryland.edu/~jalemkul
>
> ==
>
>
-- 
Gromacs Users mailing list

* Please search the archive at 
http://www.gromacs.org/Support/Mailing_Lists/GMX-Users_List before posting!

* Can't post? Read http://www.gromacs.org/Support/Mailing_Lists

* For (un)subscribe requests visit
https://maillist.sys.kth.se/mailman/listinfo/gromacs.org_gmx-users or send a 
mail to gmx-users-requ...@gromacs.org.

Re: [gmx-users] Computer processor and graphics card choice

[Guillem Prats Ejarque]

> Hi Guillem,

> I can answer your last question - the 10xx series does indeed work well
> with GROMACS (currently running GROMACS 2016 with a GTX1070), although I
> would be inclined to get the GTX1070 if your budget can stretch to it.

> Andrew

---

>> 80k atom MEM Benchmark System - 2fs time step - gmx2016.B2
>> GTX1070 - i7 6700k 4.00 GHz - 4 core - 42 ns/day - net cost per node (pc
>> chassis) ~ 950 EUR
>> GTX1070 - Xeon E5 2630v4 2.20 - 10 core - 58 ns/day - net cost per node (19"
>> rack) ~ 1600 EUR

>> GTX1080 will give you about the same performance but it is much more 
>> expensive, GTX1060 (6GB) is about 10% slower

> Note that this does not apply in general; here it is (most likely) the
> case because this specific simulation (with the GROMACS version used)
> is CPU-bound already with the 1070. If the GPU is "fast enough" for
> the workload, it does not help usign a faster GPU. Take another system
> (e.g. a CHARMM setup) and you might see significant speedup with the
> 1080.

> The 1060 being only marginally slower is also specific to the load
> balance in the MEM benchmark! With some other workload that's more
> GPU-bound on the 1060 + whichever CPU you were referring to, the 1070
> will give much larger simulation speedup.

> Certainly, if ns/day/buck is a major consideration (rather than high
> perf per node), the 1060 or 1070 are better value for the money!

> Cheers,

> Szilard


---


> Hi,

> CPU-wise, I think the 6700k, 6800k or 6850K would be better options.
> If you plan to have a second GPU, the 6700k and 6800k will not be
> ideal (too few PCI-E lanes to run both GPUs at full speed unless you
> get a special mobo). While the 2620v4 has 8 cores, it also has a low
> TDP, so it's unlikely those will run anything more than 2.5 GHz which
> won't be faster than six i7 cores at 3.3-3.5 GHz.

> If you don't need to buy right away, you might also consider the
> upcoming AMD CPUs which will be announced soon. Performance could end
> up being pretty good.

> The GTX 10xx series GPUs indeed work very well. Given the high clock
> the 1060 will certainly be faster than the 970. Andrew is right, the
> 1070 will give a further, significant bump in speed (>1.4x).

> (Note that these performance comparisons refer to the
> GPU-only/nonbonded kernel performance!)

> Cheers,

> Szilard


---

> Hi Guille,

> I'm not an expert on this, but it'll also depend on what kind of MD
> simulations you want to run. Coarse-grained forcefields, for example,
> don't benefit much from a GPU.

> So what kind of simulations do you have in mind? Atomistic? QM/MM? How
> many particles? ...?

> Peter


Thank you all for your valuable comments. Looking at that, the Xeon one really 
does not seem to be worth it. Regarding the other processors that you suggest, 
i7-6800K seems interesting to me. I do not plan to use more than one GPU.  
About the graphic card, probably I will choose the GTX1060(6GB), because 
GTX1070 is getting out of my budget.


Mostly, what I want to simulate is a protein-ligand complex on an AMBER 
forcefield, and also some QM/MM simulations. My system is small, between 30 to 
50 thousand atoms. I don't know if it is more CPU or GPU-bound.


Thanks again,


Guillem
-- 
Gromacs Users mailing list

* Please search the archive at 
http://www.gromacs.org/Support/Mailing_Lists/GMX-Users_List before posting!

* Can't post? Read http://www.gromacs.org/Support/Mailing_Lists

* For (un)subscribe requests visit
https://maillist.sys.kth.se/mailman/listinfo/gromacs.org_gmx-users or send a 
mail to gmx-users-requ...@gromacs.org.

Re: [gmx-users] trjconv sampling according to a list of arbitrary times

[Christopher Neale]

 I don't know how to do exactly what you want to do, but you can at least do 
this so that trjconv doesn't seek from the beginning of the file, it should 
speed things up a lot:

for t in 1 20 220 620; do
  let b=$t-10
  gmx trjconv -dump $t -b $b ...
done

Anyway, Mark's suggestion is much better because it will help you to avoid your 
cluster admins getting mad for you creating millions of small files ;)


From: gromacs.org_gmx-users-boun...@maillist.sys.kth.se 
 on behalf of Irem Altan 

Sent: 02 December 2016 18:00:50
To: gmx-us...@gromacs.org
Subject: [gmx-users] trjconv sampling according to a list of arbitrary times

Hi,

I have a very long simulation and I want to run trjconv on the output to save 
snapshots from a list of times. For instance if t = [1,20, 220, 620] in ps, I 
would do:

gmx trjconv -f all.trr -s nvt.tpr -dump 1 -o conf1.pdb
gmx trjconv -f all.trr -s nvt.tpr -dump 20 -o conf2.pdb
gmx trjconv -f all.trr -s nvt.tpr -dump 220 -o conf3.pdb
gmx trjconv -f all.trr -s nvt.tpr -dump 620 -o conf4.pdb

and then join them with trjcat. This is very easy to do with a script. My 
problem is, each time I run trjconv, gromacs scans the .trr file from the first 
frame until the desired frame. So if I want to save 300 such configurations, I 
have to wait for the file to be read 300 times.

Is it possible to do this with a single trjconv command, and somehow provide a 
list of the times I want to sample, so that gromacs reads the file once and 
saves the necessary snapshots?

Best,
Irem
--
Gromacs Users mailing list

* Please search the archive at 
http://www.gromacs.org/Support/Mailing_Lists/GMX-Users_List before posting!

* Can't post? Read http://www.gromacs.org/Support/Mailing_Lists

* For (un)subscribe requests visit
https://maillist.sys.kth.se/mailman/listinfo/gromacs.org_gmx-users or send a 
mail to gmx-users-requ...@gromacs.org.
-- 
Gromacs Users mailing list

* Please search the archive at 
http://www.gromacs.org/Support/Mailing_Lists/GMX-Users_List before posting!

* Can't post? Read http://www.gromacs.org/Support/Mailing_Lists

* For (un)subscribe requests visit
https://maillist.sys.kth.se/mailman/listinfo/gromacs.org_gmx-users or send a 
mail to gmx-users-requ...@gromacs.org.

Re: [gmx-users] trjconv sampling according to a list of arbitrary times

[Mark Abraham]

Hi,

There are some other ways to subsample trajectory files, check out the help
text. IIRC you can give it an index file of frame numbers, or something
like that.

Mark

On Fri, 2 Dec 2016 23:01 Irem Altan  wrote:

> Hi,
>
> I have a very long simulation and I want to run trjconv on the output to
> save snapshots from a list of times. For instance if t = [1,20, 220, 620]
> in ps, I would do:
>
> gmx trjconv -f all.trr -s nvt.tpr -dump 1 -o conf1.pdb
> gmx trjconv -f all.trr -s nvt.tpr -dump 20 -o conf2.pdb
> gmx trjconv -f all.trr -s nvt.tpr -dump 220 -o conf3.pdb
> gmx trjconv -f all.trr -s nvt.tpr -dump 620 -o conf4.pdb
>
> and then join them with trjcat. This is very easy to do with a script. My
> problem is, each time I run trjconv, gromacs scans the .trr file from the
> first frame until the desired frame. So if I want to save 300 such
> configurations, I have to wait for the file to be read 300 times.
>
> Is it possible to do this with a single trjconv command, and somehow
> provide a list of the times I want to sample, so that gromacs reads the
> file once and saves the necessary snapshots?
>
> Best,
> Irem
> --
> Gromacs Users mailing list
>
> * Please search the archive at
> http://www.gromacs.org/Support/Mailing_Lists/GMX-Users_List before
> posting!
>
> * Can't post? Read http://www.gromacs.org/Support/Mailing_Lists
>
> * For (un)subscribe requests visit
> https://maillist.sys.kth.se/mailman/listinfo/gromacs.org_gmx-users or
> send a mail to gmx-users-requ...@gromacs.org.
>
-- 
Gromacs Users mailing list

* Please search the archive at 
http://www.gromacs.org/Support/Mailing_Lists/GMX-Users_List before posting!

* Can't post? Read http://www.gromacs.org/Support/Mailing_Lists

* For (un)subscribe requests visit
https://maillist.sys.kth.se/mailman/listinfo/gromacs.org_gmx-users or send a 
mail to gmx-users-requ...@gromacs.org.

[gmx-users] trjconv sampling according to a list of arbitrary times

[Irem Altan]

Hi,

I have a very long simulation and I want to run trjconv on the output to save 
snapshots from a list of times. For instance if t = [1,20, 220, 620] in ps, I 
would do:

gmx trjconv -f all.trr -s nvt.tpr -dump 1 -o conf1.pdb
gmx trjconv -f all.trr -s nvt.tpr -dump 20 -o conf2.pdb
gmx trjconv -f all.trr -s nvt.tpr -dump 220 -o conf3.pdb
gmx trjconv -f all.trr -s nvt.tpr -dump 620 -o conf4.pdb

and then join them with trjcat. This is very easy to do with a script. My 
problem is, each time I run trjconv, gromacs scans the .trr file from the first 
frame until the desired frame. So if I want to save 300 such configurations, I 
have to wait for the file to be read 300 times.

Is it possible to do this with a single trjconv command, and somehow provide a 
list of the times I want to sample, so that gromacs reads the file once and 
saves the necessary snapshots?

Best,
Irem
-- 
Gromacs Users mailing list

* Please search the archive at 
http://www.gromacs.org/Support/Mailing_Lists/GMX-Users_List before posting!

* Can't post? Read http://www.gromacs.org/Support/Mailing_Lists

* For (un)subscribe requests visit
https://maillist.sys.kth.se/mailman/listinfo/gromacs.org_gmx-users or send a 
mail to gmx-users-requ...@gromacs.org.

Re: [gmx-users] Character issues

[Mark Abraham]

Hi,

What makes you think "_" is more relevant than "20s"?

You could well have a better experience if you were not using a version of
software that is five years old.

Matk

On Fri, 2 Dec 2016 22:45 Kelsy Yuan  wrote:

> Hi,
>
> I am trying to do an energy minimization of my system that contains stearic
> acid molecules and an ionic surface. Every time I grompp I get the
> following error massage:
>
> Program grompp_s_mpi, VERSION 4.5.5
> Source code file: string2.c, line: 105
>
> Fatal error:
> An input file contains a line longer than 4095 characters, while the buffer
> passed to fgets2 has size 4095. The line starts with: '20s'
> For more information and tips for troubleshooting, please check the GROMACS
> website at http://www.gromacs.org/Documentation/Errors
>
>
> I checked every single input file as well as the force filed directory,
> none of them contains "_" when I viewed them on different text editors. I
> wonder does anyone know another solution or possible reasons for this
> issue?
>
> Thank you in advance,
> Kelsy
> --
> Gromacs Users mailing list
>
> * Please search the archive at
> http://www.gromacs.org/Support/Mailing_Lists/GMX-Users_List before
> posting!
>
> * Can't post? Read http://www.gromacs.org/Support/Mailing_Lists
>
> * For (un)subscribe requests visit
> https://maillist.sys.kth.se/mailman/listinfo/gromacs.org_gmx-users or
> send a mail to gmx-users-requ...@gromacs.org.
>
-- 
Gromacs Users mailing list

* Please search the archive at 
http://www.gromacs.org/Support/Mailing_Lists/GMX-Users_List before posting!

* Can't post? Read http://www.gromacs.org/Support/Mailing_Lists

* For (un)subscribe requests visit
https://maillist.sys.kth.se/mailman/listinfo/gromacs.org_gmx-users or send a 
mail to gmx-users-requ...@gromacs.org.

[gmx-users] Character issues

[Kelsy Yuan]

Hi,

I am trying to do an energy minimization of my system that contains stearic
acid molecules and an ionic surface. Every time I grompp I get the
following error massage:

Program grompp_s_mpi, VERSION 4.5.5
Source code file: string2.c, line: 105

Fatal error:
An input file contains a line longer than 4095 characters, while the buffer
passed to fgets2 has size 4095. The line starts with: '20s'
For more information and tips for troubleshooting, please check the GROMACS
website at http://www.gromacs.org/Documentation/Errors


I checked every single input file as well as the force filed directory,
none of them contains "_" when I viewed them on different text editors. I
wonder does anyone know another solution or possible reasons for this issue?

Thank you in advance,
Kelsy
-- 
Gromacs Users mailing list

* Please search the archive at 
http://www.gromacs.org/Support/Mailing_Lists/GMX-Users_List before posting!

* Can't post? Read http://www.gromacs.org/Support/Mailing_Lists

* For (un)subscribe requests visit
https://maillist.sys.kth.se/mailman/listinfo/gromacs.org_gmx-users or send a 
mail to gmx-users-requ...@gromacs.org.

Re: [gmx-users] PMF decompostion

[Christopher Neale]

PMF gives free energy, which can not be decomposed into VDW and Q without 
leaving a remainder, though I guess that doesn't affect your question. Enthalpy 
is simply the average potential energy plus the pressure volume component. So 
you can compute average potential energy VDW and Q components at each umbrella 
if you want (from the .edr files with gmx energy). Be aware that the 
statistical uncertainty on the enthalpy tends to be greater than the relevant 
differences on achievable simulation timescales (i.e., convergence is hard), so 
you'll want to evaluate error explicitly.


From: gromacs.org_gmx-users-boun...@maillist.sys.kth.se 
 on behalf of Salman Zarrini 

Sent: 02 December 2016 17:16:15
To: gmx-us...@gromacs.org
Subject: [gmx-users] PMF decompostion

Dear gromacs user,

I was wondering if something has been already implemented in gromacs to
make possible the potential of mean force (PMF) decomposition into the van
der waals and electrostatic energies?

Would you please also let me know if the desolvation energy originated from
protein-water or in my case solid surface-water interaction, will be taken
into account automatically when one calculates the PMF in aqueous solution
in gromacs?

Thanks.

Cheers,
Salman Zarrini
--
Gromacs Users mailing list

* Please search the archive at 
http://www.gromacs.org/Support/Mailing_Lists/GMX-Users_List before posting!

* Can't post? Read http://www.gromacs.org/Support/Mailing_Lists

* For (un)subscribe requests visit
https://maillist.sys.kth.se/mailman/listinfo/gromacs.org_gmx-users or send a 
mail to gmx-users-requ...@gromacs.org.
-- 
Gromacs Users mailing list

* Please search the archive at 
http://www.gromacs.org/Support/Mailing_Lists/GMX-Users_List before posting!

* Can't post? Read http://www.gromacs.org/Support/Mailing_Lists

* For (un)subscribe requests visit
https://maillist.sys.kth.se/mailman/listinfo/gromacs.org_gmx-users or send a 
mail to gmx-users-requ...@gromacs.org.

[gmx-users] PMF decompostion

[Salman Zarrini]

Dear gromacs user,

I was wondering if something has been already implemented in gromacs to
make possible the potential of mean force (PMF) decomposition into the van
der waals and electrostatic energies?

Would you please also let me know if the desolvation energy originated from
protein-water or in my case solid surface-water interaction, will be taken
into account automatically when one calculates the PMF in aqueous solution
in gromacs?

Thanks.

Cheers,
Salman Zarrini
-- 
Gromacs Users mailing list

* Please search the archive at 
http://www.gromacs.org/Support/Mailing_Lists/GMX-Users_List before posting!

* Can't post? Read http://www.gromacs.org/Support/Mailing_Lists

* For (un)subscribe requests visit
https://maillist.sys.kth.se/mailman/listinfo/gromacs.org_gmx-users or send a 
mail to gmx-users-requ...@gromacs.org.

Re: [gmx-users] Computer processor and graphics card choice

[Szilárd Páll]

On Fri, Dec 2, 2016 at 1:36 PM, T.M.D. Graen  wrote:
> 80k atom MEM Benchmark System - 2fs time step - gmx2016.B2
> GTX1070 - i7 6700k 4.00 GHz - 4 core - 42 ns/day - net cost per node (pc
> chassis) ~ 950 EUR
> GTX1070 - Xeon E5 2630v4 2.20 - 10 core - 58 ns/day - net cost per node (19"
> rack) ~ 1600 EUR
>

> GTX1080 will give you about the same performance but it is much more 
> expensive, GTX1060 (6GB) is about 10% slower

Note that this does not apply in general; here it is (most likely) the
case because this specific simulation (with the GROMACS version used)
is CPU-bound already with the 1070. If the GPU is "fast enough" for
the workload, it does not help usign a faster GPU. Take another system
(e.g. a CHARMM setup) and you might see significant speedup with the
1080.

The 1060 being only marginally slower is also specific to the load
balance in the MEM benchmark! With some other workload that's more
GPU-bound on the 1060 + whichever CPU you were referring to, the 1070
will give much larger simulation speedup.

Certainly, if ns/day/buck is a major consideration (rather than high
perf per node), the 1060 or 1070 are better value for the money!

Cheers,
--
Szilárd

>
>
> On 12/01/2016 08:15 PM, Guillem Prats Ejarque wrote:
>>
>> Dear colleagues,
>>
>>
>> I want to buy a computer to perform molecular dynamics by GROMACS, and I
>> have several questions.
>>
>>
>> First of all, I am having doubts between the Intel Xeon E5-2620v4 (8
>> cores, 2.1-3.0 GHz) and the i7-5820K (6 cores, 3.3-3.6 GHz). The reason of
>> my doubts is the significantly lower clock rate of the Xeon processor. First
>> I decided i7-5820k, but, although E5-2620 is at the limit of my budget, I
>> would buy it only if the performance were significantly better.
>>
>>
>> Moreover, after the release of the new 10xx NVIDIA series, I also have
>> doubts about the graphics card. First I wanted to buy the GTX 970 (1664 CUDA
>> cores; 1050-1178 MHz clock; 224 GB/s Memory bandwidth; 3494 SP Gflop/s), but
>> looking at the new GTX 1060 - 6 GB (1280 CUDA cores; 1506-1708 MHz clock;
>> 192 GB/s Memory bandwidth; 3855 SP Gflop/s) which is in the same price
>> range,  it seemed to me that it could have higher performance with the last
>> one. Would GTX 1060 be better than GTX 970? Does anyone know if the 10xx
>> series of NVIDA works well with GROMACS?
>>
>>
>> Thanks in advance,
>>
>>
>> Guillem
>>
>
> --
> Gromacs Users mailing list
>
> * Please search the archive at
> http://www.gromacs.org/Support/Mailing_Lists/GMX-Users_List before posting!
>
> * Can't post? Read http://www.gromacs.org/Support/Mailing_Lists
>
> * For (un)subscribe requests visit
> https://maillist.sys.kth.se/mailman/listinfo/gromacs.org_gmx-users or send a
> mail to gmx-users-requ...@gromacs.org.
-- 
Gromacs Users mailing list

* Please search the archive at 
http://www.gromacs.org/Support/Mailing_Lists/GMX-Users_List before posting!

* Can't post? Read http://www.gromacs.org/Support/Mailing_Lists

* For (un)subscribe requests visit
https://maillist.sys.kth.se/mailman/listinfo/gromacs.org_gmx-users or send a 
mail to gmx-users-requ...@gromacs.org.

Re: [gmx-users] Computer processor and graphics card choice

[Szilárd Páll]

Hi,

CPU-wise, I think the 6700k, 6800k or 6850K would be better options.
If you plan to have a second GPU, the 6700k and 6800k will not be
ideal (too few PCI-E lanes to run both GPUs at full speed unless you
get a special mobo). While the 2620v4 has 8 cores, it also has a low
TDP, so it's unlikely those will run anything more than 2.5 GHz which
won't be faster than six i7 cores at 3.3-3.5 GHz.

If you don't need to buy right away, you might also consider the
upcoming AMD CPUs which will be announced soon. Performance could end
up being pretty good.

The GTX 10xx series GPUs indeed work very well. Given the high clock
the 1060 will certainly be faster than the 970. Andrew is right, the
1070 will give a further, significant bump in speed (>1.4x).

(Note that these performance comparisons refer to the
GPU-only/nonbonded kernel performance!)

Cheers,
--
Szilárd


On Thu, Dec 1, 2016 at 8:15 PM, Guillem Prats Ejarque
 wrote:
> Dear colleagues,
>
>
> I want to buy a computer to perform molecular dynamics by GROMACS, and I have 
> several questions.
>
>
> First of all, I am having doubts between the Intel Xeon E5-2620v4 (8 cores, 
> 2.1-3.0 GHz) and the i7-5820K (6 cores, 3.3-3.6 GHz). The reason of my doubts 
> is the significantly lower clock rate of the Xeon processor. First I decided 
> i7-5820k, but, although E5-2620 is at the limit of my budget, I would buy it 
> only if the performance were significantly better.
>
>
> Moreover, after the release of the new 10xx NVIDIA series, I also have doubts 
> about the graphics card. First I wanted to buy the GTX 970 (1664 CUDA cores; 
> 1050-1178 MHz clock; 224 GB/s Memory bandwidth; 3494 SP Gflop/s), but looking 
> at the new GTX 1060 - 6 GB (1280 CUDA cores; 1506-1708 MHz clock; 192 GB/s 
> Memory bandwidth; 3855 SP Gflop/s) which is in the same price range,  it 
> seemed to me that it could have higher performance with the last one. Would 
> GTX 1060 be better than GTX 970? Does anyone know if the 10xx series of NVIDA 
> works well with GROMACS?
>
>
> Thanks in advance,
>
>
> Guillem
> --
> Gromacs Users mailing list
>
> * Please search the archive at 
> http://www.gromacs.org/Support/Mailing_Lists/GMX-Users_List before posting!
>
> * Can't post? Read http://www.gromacs.org/Support/Mailing_Lists
>
> * For (un)subscribe requests visit
> https://maillist.sys.kth.se/mailman/listinfo/gromacs.org_gmx-users or send a 
> mail to gmx-users-requ...@gromacs.org.
-- 
Gromacs Users mailing list

* Please search the archive at 
http://www.gromacs.org/Support/Mailing_Lists/GMX-Users_List before posting!

* Can't post? Read http://www.gromacs.org/Support/Mailing_Lists

* For (un)subscribe requests visit
https://maillist.sys.kth.se/mailman/listinfo/gromacs.org_gmx-users or send a 
mail to gmx-users-requ...@gromacs.org.

Re: [gmx-users] Removal of Waters in Hydrophobic Core

[Christopher Neale]

Overall:
A) you have an incorrect new_waters.gro file
B) your dos formatting is confusing the script
C) your definitions for Z-based removal make no sense
D) your input file (pre-water) is, at the very least, unique

More details:

First, why are your lipids so sparse in X/Y? This looks like a file that would 
be used for the start of inflategro but has not been recompacted yet. It's fine 
if this is really what you want, but I don't see why you would.

As to your problems running the script to remove waters:
1. new_waters.gro should not be an actual .gro file should only have atom 
definitions, no header or tail for the .gro format. I realize that it is 
confusing that I called it new_waters.gro ... sorry.
2. your files don't work for many commands since they are in dos format. I used 
dos2unix to fix some (like scripts), but even that one did not completely heal 
initial.gro, which still reported 6440 lines (should be 6441) from the linux 
tool "wc". I fixed with gmx editconf.
3. I'm not sure what you want to do gro file units are nm, and your box is 
6 nm tall but you want to remove a range from 5 to 50 nm?
4. anyway the C program on the same page ( 
http://www.gromacs.org/Documentation/How-tos/Membrane_Simulations ) is *way* 
faster, so why not use that instead? I tested it with your inputs and, after 
making fixes that I mentioned to have a correct new_water.gro input and get rid 
of dos-formatting problems, it worked as expected. -- I did not verify the bash 
script version since it takes forever to compute with such an enormous system. 
I suggest if you are going to stick with that script that you start by 
debugging with a *way* smaller system first.


From: gromacs.org_gmx-users-boun...@maillist.sys.kth.se 
 on behalf of Sanim Rahman 

Sent: 29 November 2016 02:30:35
To: gmx-us...@gromacs.org
Subject: Re: [gmx-users] Removal of Waters in Hydrophobic Core

Thank you for the responses.

Here is a Dropbox link to all the files and scripts I am using:

https://www.dropbox.com/sh/tk7syifmtwqmtn0/AABWwtX5XNzzBiJB08kQ-xmca?dl=0

As I mentioned before, I am working through the second tutorial of the
Bevan Labs GROMACS Tutorial (KALP15 in DPPC). I am attempting to remove
waters from the hydrophobic core.

I attempted to remove waters from the hydrophobic core by
specifying z-coordinates since the system is oriented along the z-axis. I
experimented with an extreme range in the script and images I uploaded
(lowerz= 5.000 upperz= 50.000).

Thanks for the assistance! I deeply appreciate the time you have spent with
my situation.

Regards,

*Sanim Rahman*
B.S. Chemical Engineering, 2019
Undergraduate Researcher, Global Center for Hearing and Speech Research



On Mon, Nov 28, 2016 at 9:11 PM, Justin Lemkul  wrote:

>
>
> On 11/28/16 7:02 PM, Christopher Neale wrote:
>
>> You might get more help if you post your script, .gro file, images, etc.
>> somewhere online and link them in a mailing list message. At this point
>> it's impossible to know what is going wrong for your usage with the
>> information you have posted.
>>
>> I don't know about Justin's scripts. Perhaps they could work together,
>> but you have not said what you want to do.
>>
>> An alternative is simply to use good selections with "gmx select" to get
>> rid of the waters inside your bilayer.
>>
>>
> My script might be useful in this case, but only with groups derived from
> what it provides.  It prints index groups of the top and bottom leaflet
> lipids.  gmx traj -ox -com can then be used on a suitable group's P atoms
> (or some other atoms within, e.g. the glycerol moiety) to determine the
> upper and lower coordinate values defining the boundaries of the membrane
> for removal of water.
>
> But gmx select would also be a good option. Probably faster, too.
>
> -Justin
>
>
> 
>> From: gromacs.org_gmx-users-boun...@maillist.sys.kth.se <
>> gromacs.org_gmx-users-boun...@maillist.sys.kth.se> on behalf of Sanim
>> Rahman 
>> Sent: 28 November 2016 14:26:07
>> To: gmx-us...@gromacs.org
>> Subject: Re: [gmx-users] Removal of Waters in Hydrophobic Core
>>
>> Hey Chris,
>>
>> Your instruction makes much more sense. I was really confused on what was
>> the point of generating the keep_these_waters.gro file initially. Thank
>> you. I have noticed that water molecules were removed from my structure,
>> however, it removed waters I was not looking for. It took off a random
>> column of water molecules in my bilayer. What type of data file (gro, top,
>> pdb, etc.) or graphical interface program is recommended to decide my
>> upperz and lowerz value?
>>
>> I know that I am now jumping around with my questions, but are you
>> familiar
>> with the scripts on the site below? I believe that the
>> 

Re: [gmx-users] gmx distance -- gmx: double free or corruption

[Christopher Neale]

Posted:
https://redmine.gromacs.org/issues/2086

Thanks Mark.

From: gromacs.org_gmx-users-boun...@maillist.sys.kth.se 
 on behalf of Mark Abraham 

Sent: 29 November 2016 09:27:34
To: Discussion list for GROMACS users
Subject: Re: [gmx-users] gmx distance -- gmx: double free or corruption

Hi,

That looks like something can be improved in the code, but it's hard to say
what. Can you open another Redmine issue with repo files please? If a
single pdb frame reproduces it, then that would likely be best.

I see you've tried to use -sf to simplify common subexpressions - that's
likely a good idea in general. See e.g. gmx help selections evaluation.

Mark

On Fri, Nov 25, 2016 at 8:45 PM Christopher Neale <
chris.ne...@alum.utoronto.ca> wrote:

> Please ignore the mention of this being an amber trajectory. I can
> reproduce the exact same behaviour by gmx distance -s my.pdb -f my.pdb , so
> the issue can be reproduced on the information in a PDB file alone (and
> that pdb looks just fine in VMD).
>
> Thank you,
> Chris.
>
> 
> From: gromacs.org_gmx-users-boun...@maillist.sys.kth.se <
> gromacs.org_gmx-users-boun...@maillist.sys.kth.se> on behalf of
> Christopher Neale 
> Sent: 25 November 2016 14:33:51
> To: Discussion list for GROMACS users
> Subject: [gmx-users] gmx distance -- gmx: double free or corruption
>
> This sender failed our fraud detection checks and may not be who they
> appear to be. Learn about spoofing at http://aka.ms/LearnAboutSpoofing
>
> Dear Users:
>
> I am having trouble with gmx distance crashing in gromacs 5.1.2. I am
> trying to analyze the distance of different peptide residues from the
> center of a lipid bilayer. have CA atoms on residues 2-20 and also on
> residues 22-40 (2 peptides). If I try to do residues 2-20 it works fine. If
> I try to do residues 22 to 40, I get a memory error. Funny thing is, I
> narrowed it down to see that 22-36 is OK but 22-37 fails, then when I try
> 37 alone it is OK, and selections of 26-40 also work. Please note that all
> these tests were repeated 2x on an allocation with 128 GB of memory, so I
> don’t think I’m just describing a stochastic failure.
>
> I didn’t try newer versions of gromacs, but there is nothing in the
> release notes for any newer version that mentions “distance” or “select”
> that seems to be related to this.
>
> ## I run like this:
> inpstr=$(for((i=22;i<=40;i++)); do echo -n "com of (residue $i and name
> CA) plus com of (resname POPC or resname POPS) plus "; done|sed "s/plus
> $//")
> gmx distance -s gromacs.tpr -f gromacs.xtc -n index.ndx -select "$inpstr"
> -oxyz depthB.xvg
>
> I’ve also tried giving gmx distance a .pdb as input to the -s option,
> where I first ran gmx editconf -resnr 1 just to avoid any issues with
> possible double residue counting in the .tpr file, and the result is the
> same crash.
>
> I checked that the .pdb and the .xtc have the same number of atoms with
> gmx check  and I also ran gmx select on these groups to ensure that an atom
> was picked up in all cases (it was). Also, VMD loading the .pdb extracted
> from the .tpr and the .xtc gives a sensible trajectory.
>
> As a final note, the simulation was actually run in amber. I then used a
> plugin-aware version go gmx trjconv to read in the amber netcdf trajectory
> file and output a gromacs .xtc file, which I used in a non-plugin-aware
> version of gromacs for the analyses. The .tpr was generated for the same
> system (actually, I generated the gromacs .tpr from a grompp -pp .top file
> that was used in amber tool parmed gromber to set up the amber run).
>
> ## The output (with error) is this:
>
>  :-) GROMACS - gmx distance, VERSION 5.1.2 (-:
>
> GROMACS is written by:
>  Emile Apol  Rossen Apostolov  Herman J.C. BerendsenPar
> Bjelkmar
>  Aldert van Buuren   Rudi van Drunen Anton Feenstra   Sebastian Fritsch
>   Gerrit Groenhof   Christoph Junghans   Anca HamuraruVincent
> Hindriksen
>  Dimitrios KarkoulisPeter KassonJiri Kraus  Carsten Kutzner
> Per Larsson  Justin A. Lemkul   Magnus Lundborg   Pieter Meulenhoff
>Erik Marklund  Teemu Murtola   Szilard Pall   Sander Pronk
>Roland Schulz Alexey Shvetsov Michael Shirts Alfons Sijbers
>Peter TielemanTeemu Virolainen  Christian WennbergMaarten Wolf
>and the project leaders:
> Mark Abraham, Berk Hess, Erik Lindahl, and David van der Spoel
>
> Copyright (c) 1991-2000, University of Groningen, The Netherlands.
> Copyright (c) 2001-2015, The GROMACS development team at
> Uppsala University, Stockholm University and
> the Royal Institute of Technology, Sweden.
> check out http://www.gromacs.org for more information.
>
> GROMACS is free software; you can 

Re: [gmx-users] COM group of Membrane and Protein simulation

[Mijiddorj Batsaikhan]

Dear Justin,

Thank you very much.
I attached my mdp file. Peptide locates on membrane surface. I want to know
about interaction between membrane and peptide, membrane permeability of
peptide, amino acids contribution for the penetration, specially energetic
values between the membrane and the peptide.
Please suggest me with helpful options?
How can I chose COM groups?


--

>
> Message: 5
> Date: Fri, 2 Dec 2016 07:50:52 -0500
> From: Justin Lemkul 
> To: gmx-us...@gromacs.org
> Subject: Re: [gmx-users] COM group of Membrane and Protein simulation
> Message-ID: 
> Content-Type: text/plain; charset=windows-1252; format=flowed
>
>
>
> On 12/1/16 9:17 PM, Mijiddorj Batsaikhan wrote:
> > Dear gmx_users,
> >
> > I started simulation that a peptide on membrane. My peptide locates on
> the
> > membrane surface. I have two questions relating to the simulation.
> > (1)
> > When I start the simulation, I chose COM groups separately. Is this
> choice
> > okay? or May I need to chose COM group inseparately?
> >
>
> Your description is ambiguous; please post the actual .mdp contents.
>
> > (2)
> > During the simulation peptide is moving the edge of membrane. How can I
> > shift the peptide to the central section of the membrane? Can I use
> > -nojump, -center options of trjconv tool?
> >
>
> Yes - try it and see.  Also relevant is -fit transxy.
>
> -Justin
>
> --
> ==
>
> Justin A. Lemkul, Ph.D.
> Ruth L. Kirschstein NRSA Postdoctoral Fellow
>
> Department of Pharmaceutical Sciences
> School of Pharmacy
> Health Sciences Facility II, Room 629
> University of Maryland, Baltimore
> 20 Penn St.
> Baltimore, MD 21201
>
> jalem...@outerbanks.umaryland.edu | (410) 706-7441
> http://mackerell.umaryland.edu/~jalemkul
>
> ==
>
>
-- 
Gromacs Users mailing list

* Please search the archive at 
http://www.gromacs.org/Support/Mailing_Lists/GMX-Users_List before posting!

* Can't post? Read http://www.gromacs.org/Support/Mailing_Lists

* For (un)subscribe requests visit
https://maillist.sys.kth.se/mailman/listinfo/gromacs.org_gmx-users or send a 
mail to gmx-users-requ...@gromacs.org.

[gmx-users] About negative values deuterium order parameters

[Poncho Arvayo Zatarain]

Some days ago i send a quetion about negative values of deuterium order 
parameters in a 25% DPPC 75% DPPE+ nesthesic membrane. Thomas Piggot repond to 
me about the Scd que be negative too, specially for DPPC and DPPE. Now i send 
my obatained values for Scd. It is correct to have negative values? SN-1 and 
SN-3 correspond to DPPC and SN-2 and SN-4 correspond to DPPE. I watch the gro 
file in vmd and certainly look inclinable for the tails. All my other systems 
don´t have negative values like this one.

Carbon  DPPC64+DPPE192+AN-SN1   DPPC64+DPPE192+AN-SN2   DPPC64+DPPE192+AN-SN3   
DPPC64+DPPE192+AN-SN4
1   -0.207868   0.156621-0.122543   0.202863
2   -0.199115   0.0507337   0.0757880.0911758
3   -0.129733   0.0328502   0.0638192   0.0722099
4   0.159087-0.0813808  0.0432806   0.0571403
5   -0.0940551  -0.0807225  -0.0842683  0.189616
6   -0.199373   -0.0259852  -0.0721477  0.167733
7   0.122858-0.0486781  0.00461267  0.132131
8   0.135019-0.0244167  -0.0931847  0.132473
9   0.1342240.0127981   0.0474575   0.143573
10  0.1350230.0502274   0.0901911   0.132179
11  0.128295-0.0841136  0.0882197   0.0943604
12  0.1252270.0756042   0.0349276   0.12459
13  0.1159380.220139-0.0766258  0.0861008

-- 
Gromacs Users mailing list

* Please search the archive at 
http://www.gromacs.org/Support/Mailing_Lists/GMX-Users_List before posting!

* Can't post? Read http://www.gromacs.org/Support/Mailing_Lists

* For (un)subscribe requests visit
https://maillist.sys.kth.se/mailman/listinfo/gromacs.org_gmx-users or send a 
mail to gmx-users-requ...@gromacs.org.

Re: [gmx-users] Using GROMACS with (AMD) OpenCL

[Szilárd Páll]

Hi,

While for the pre-Polaris hardware Milan is right, for Polaris you
*need* a AMDGPU-PRO (which only runs on newer distros/kernels).

The current AMDGPU-PRO from the OpenCL point of view is, AFAIK more or
less the same compiler/runtime as the last catalyst release (15.12).

One notable caveat is that with AMDGPU-PRO power management seems to
be overly aggressive (messed up?) -- at least for the load GROMACS
generates. Uou can loose quite some performance unless you force the
performance level to "high" (hint use radcard to set
power_dpm_force_performance_level=high); this has been obsered on both
Polaris and Fiji.

See here:
https://wiki.archlinux.org/index.php/ATI#Dynamic_power_management

Cheers,
--
Szilárd


On Fri, Dec 2, 2016 at 4:13 PM,   wrote:
> Hi Gregory,
> it is possible to use with both - Catalyst or AMDGPU, so you can choose. As 
> there won't be (I think) any more updates for Catalyst, I'd prefer to use 
> newer AMDGPU (PRO) choice. I don't have Polaris based card, only some 79xx/R9 
> 280 (X) (uding Catalyst) and one Fury Nano (used Catalyst also, but now 
> AMDGPU PRO). I made the Gromacs' compilation only with the Catalyst drivers, 
> but with AMDGPU shouldn't be more complicated. Just install the system with 
> AMDGPU (I don't know it "PRO part" is needed, probably yes) and try to 
> compile (don't forget about the OpenCL switch). If there would be some 
> missing libraries, than you have to install some -dev versions of packages. 
> Probably Ubuntu has it's own way how to find the right package, but you can 
> also use site at packages.debian.org (as Ubuntu is based on Debian).
> Anyway I'm trying to have most of things directly from the distribution, 
> cause it would automatically upgrades and fixes bugs. Outside the distro I 
> have only Gromacs (performancd issues - compilation to exact machine - and 
> several version installed) and AMDGPU PRO (on that node with Fury Nano, cause 
> Debian doesn't have it in its repositories.
> If any problem, plaase ask more with some details.
>
> Best,
>
> Milan
>
> On Fri, Dec 02, 2016 at 01:56:03PM +, Gregory Man Kai Poon wrote:
>> Hello all,
>>
>>
>> I would like to get some practical advice on setting up GROMACS with OpenCL 
>> using an AMD GPU (an RX 480 is what I happen to have).  The AMD APP SDK 
>> indicates that it supports the Catalyst Omega 15.7 driver -- is that 
>> compatible to the proprietary Catalyst driver for Linux?  Since this 
>> proprietary driver is only compatible with Ubuntu up to version 15.10, does 
>> that mean that best software configuration (say for GROMACS 5.1.4) is:
>>
>>
>> Ubuntu 15.10 + propriety Catalyst driver + AMD APP SDK
>>
>>
>> There is also a newer AMDGPU driver that works with Ubuntu 16.04 -- is that 
>> compatible with the AMD APP SDK and would allow GROMACS to be set up on a 
>> newer OS?  If I really should be going about some other way (given the GPU), 
>> please suggest also.
>>
>>
>> I have googled the subject a bit but was unable to find enough info to 
>> proceed.  Thanks in advance for your insight.
>>
>>
>> Best regards,
>>
>> Gregory
>> --
>> Gromacs Users mailing list
>>
>> * Please search the archive at 
>> http://www.gromacs.org/Support/Mailing_Lists/GMX-Users_List before posting!
>>
>> * Can't post? Read http://www.gromacs.org/Support/Mailing_Lists
>>
>> * For (un)subscribe requests visit
>> https://maillist.sys.kth.se/mailman/listinfo/gromacs.org_gmx-users or send a 
>> mail to gmx-users-requ...@gromacs.org.
> --
> Gromacs Users mailing list
>
> * Please search the archive at 
> http://www.gromacs.org/Support/Mailing_Lists/GMX-Users_List before posting!
>
> * Can't post? Read http://www.gromacs.org/Support/Mailing_Lists
>
> * For (un)subscribe requests visit
> https://maillist.sys.kth.se/mailman/listinfo/gromacs.org_gmx-users or send a 
> mail to gmx-users-requ...@gromacs.org.
-- 
Gromacs Users mailing list

* Please search the archive at 
http://www.gromacs.org/Support/Mailing_Lists/GMX-Users_List before posting!

* Can't post? Read http://www.gromacs.org/Support/Mailing_Lists

* For (un)subscribe requests visit
https://maillist.sys.kth.se/mailman/listinfo/gromacs.org_gmx-users or send a 
mail to gmx-users-requ...@gromacs.org.

Re: [gmx-users] Using GROMACS with (AMD) OpenCL

[melichercik]

Hi Gregory,
it is possible to use with both - Catalyst or AMDGPU, so you can choose. As 
there won't be (I think) any more updates for Catalyst, I'd prefer to use newer 
AMDGPU (PRO) choice. I don't have Polaris based card, only some 79xx/R9 280 (X) 
(uding Catalyst) and one Fury Nano (used Catalyst also, but now AMDGPU PRO). I 
made the Gromacs' compilation only with the Catalyst drivers, but with AMDGPU 
shouldn't be more complicated. Just install the system with AMDGPU (I don't 
know it "PRO part" is needed, probably yes) and try to compile (don't forget 
about the OpenCL switch). If there would be some missing libraries, than you 
have to install some -dev versions of packages. Probably Ubuntu has it's own 
way how to find the right package, but you can also use site at 
packages.debian.org (as Ubuntu is based on Debian).
Anyway I'm trying to have most of things directly from the distribution, cause 
it would automatically upgrades and fixes bugs. Outside the distro I have only 
Gromacs (performancd issues - compilation to exact machine - and several 
version installed) and AMDGPU PRO (on that node with Fury Nano, cause Debian 
doesn't have it in its repositories.
If any problem, plaase ask more with some details.

Best,

Milan

On Fri, Dec 02, 2016 at 01:56:03PM +, Gregory Man Kai Poon wrote:
> Hello all,
> 
> 
> I would like to get some practical advice on setting up GROMACS with OpenCL 
> using an AMD GPU (an RX 480 is what I happen to have).  The AMD APP SDK 
> indicates that it supports the Catalyst Omega 15.7 driver -- is that 
> compatible to the proprietary Catalyst driver for Linux?  Since this 
> proprietary driver is only compatible with Ubuntu up to version 15.10, does 
> that mean that best software configuration (say for GROMACS 5.1.4) is:
> 
> 
> Ubuntu 15.10 + propriety Catalyst driver + AMD APP SDK
> 
> 
> There is also a newer AMDGPU driver that works with Ubuntu 16.04 -- is that 
> compatible with the AMD APP SDK and would allow GROMACS to be set up on a 
> newer OS?  If I really should be going about some other way (given the GPU), 
> please suggest also.
> 
> 
> I have googled the subject a bit but was unable to find enough info to 
> proceed.  Thanks in advance for your insight.
> 
> 
> Best regards,
> 
> Gregory
> -- 
> Gromacs Users mailing list
> 
> * Please search the archive at 
> http://www.gromacs.org/Support/Mailing_Lists/GMX-Users_List before posting!
> 
> * Can't post? Read http://www.gromacs.org/Support/Mailing_Lists
> 
> * For (un)subscribe requests visit
> https://maillist.sys.kth.se/mailman/listinfo/gromacs.org_gmx-users or send a 
> mail to gmx-users-requ...@gromacs.org.
-- 
Gromacs Users mailing list

* Please search the archive at 
http://www.gromacs.org/Support/Mailing_Lists/GMX-Users_List before posting!

* Can't post? Read http://www.gromacs.org/Support/Mailing_Lists

* For (un)subscribe requests visit
https://maillist.sys.kth.se/mailman/listinfo/gromacs.org_gmx-users or send a 
mail to gmx-users-requ...@gromacs.org.

[gmx-users] Using GROMACS with (AMD) OpenCL

[Gregory Man Kai Poon]

Hello all,


I would like to get some practical advice on setting up GROMACS with OpenCL 
using an AMD GPU (an RX 480 is what I happen to have).  The AMD APP SDK 
indicates that it supports the Catalyst Omega 15.7 driver -- is that compatible 
to the proprietary Catalyst driver for Linux?  Since this proprietary driver is 
only compatible with Ubuntu up to version 15.10, does that mean that best 
software configuration (say for GROMACS 5.1.4) is:


Ubuntu 15.10 + propriety Catalyst driver + AMD APP SDK


There is also a newer AMDGPU driver that works with Ubuntu 16.04 -- is that 
compatible with the AMD APP SDK and would allow GROMACS to be set up on a newer 
OS?  If I really should be going about some other way (given the GPU), please 
suggest also.


I have googled the subject a bit but was unable to find enough info to proceed. 
 Thanks in advance for your insight.


Best regards,

Gregory
-- 
Gromacs Users mailing list

* Please search the archive at 
http://www.gromacs.org/Support/Mailing_Lists/GMX-Users_List before posting!

* Can't post? Read http://www.gromacs.org/Support/Mailing_Lists

* For (un)subscribe requests visit
https://maillist.sys.kth.se/mailman/listinfo/gromacs.org_gmx-users or send a 
mail to gmx-users-requ...@gromacs.org.

Re: [gmx-users] COM group of Membrane and Protein simulation

[Justin Lemkul]

On 12/1/16 9:17 PM, Mijiddorj Batsaikhan wrote:

Dear gmx_users,

I started simulation that a peptide on membrane. My peptide locates on the
membrane surface. I have two questions relating to the simulation.
(1)
When I start the simulation, I chose COM groups separately. Is this choice
okay? or May I need to chose COM group inseparately?



Your description is ambiguous; please post the actual .mdp contents.


(2)
During the simulation peptide is moving the edge of membrane. How can I
shift the peptide to the central section of the membrane? Can I use
-nojump, -center options of trjconv tool?



Yes - try it and see.  Also relevant is -fit transxy.

-Justin

--
==

Justin A. Lemkul, Ph.D.
Ruth L. Kirschstein NRSA Postdoctoral Fellow

Department of Pharmaceutical Sciences
School of Pharmacy
Health Sciences Facility II, Room 629
University of Maryland, Baltimore
20 Penn St.
Baltimore, MD 21201

jalem...@outerbanks.umaryland.edu | (410) 706-7441
http://mackerell.umaryland.edu/~jalemkul

==
--
Gromacs Users mailing list

* Please search the archive at 
http://www.gromacs.org/Support/Mailing_Lists/GMX-Users_List before posting!

* Can't post? Read http://www.gromacs.org/Support/Mailing_Lists

* For (un)subscribe requests visit
https://maillist.sys.kth.se/mailman/listinfo/gromacs.org_gmx-users or send a 
mail to gmx-users-requ...@gromacs.org.

Re: [gmx-users] system blowing up

[Justin Lemkul]

On 12/1/16 1:23 PM, abhisek Mondal wrote:

Hi,

I'm constantly getting a strange error of system blow up during
equilibration procedure. I'm relatively new to gromacs, please help me to
get my things straight.

I have a ligand bound protein structure in solvated condition.
So, I was using:
gmx mdrun -v -deffnm em #for minimization

Upon successful completion, I try to equilibrate using:
gmx mdrun -deffnm npt

But this command is failing with a lot of LINCS errors.
Surprising thing is that, I'm running the same code in 2 clusters
simultaneously. In one cluster it is running fine but in other cluster it
is giving me LINCS error.

Can you please suggest what actually is going wrong.



http://www.gromacs.org/Documentation/Terminology/Blowing_Up#Diagnosing_an_Unstable_System

-Justin

--
==

Justin A. Lemkul, Ph.D.
Ruth L. Kirschstein NRSA Postdoctoral Fellow

Department of Pharmaceutical Sciences
School of Pharmacy
Health Sciences Facility II, Room 629
University of Maryland, Baltimore
20 Penn St.
Baltimore, MD 21201

jalem...@outerbanks.umaryland.edu | (410) 706-7441
http://mackerell.umaryland.edu/~jalemkul

==
--
Gromacs Users mailing list

* Please search the archive at 
http://www.gromacs.org/Support/Mailing_Lists/GMX-Users_List before posting!

* Can't post? Read http://www.gromacs.org/Support/Mailing_Lists

* For (un)subscribe requests visit
https://maillist.sys.kth.se/mailman/listinfo/gromacs.org_gmx-users or send a 
mail to gmx-users-requ...@gromacs.org.

Re: [gmx-users] hbond

[Justin Lemkul]

On 12/1/16 10:58 AM, Gregory Poon wrote:



On 12/1/2016 7:45 AM, Justin Lemkul wrote:



On 12/1/16 7:24 AM, Gregory Man Kai Poon wrote:

Hi all:


I am trying to use hbond in Gromacs 5.1.4 to analyze water-mediated hydrogen
bonding between two objects simulated in water.  The GROMACS manual discusses
this in a Figure (9.8) - "water insertion". There is nothing in the online
documentation as to how this should be done except a single mention with the
-hbm option, which I tried.  It generated .xpm files such as the one
attached.  They open, as far as I can tell, a very vertically compressed plot
which I can make nothing out of.  Attempts to convert them to eps using
xpm2eps output similar results.



You can use an .m2p file to adjust the sizes of the x- and y-axes to make it
legible.  The real value is in the data within, though.  You have to map the
actual participating groups (the output of of -hbn) with the individual time
series in the .xpm from -hbm.



So my questions are two-fold: 1) What is happening with the .xpm files?  2)
Am I using the correct hbond option to enumerate water-mediated hydrogen
bonds?



To actually analyze water-mediated H-bonds requires additional work that
GROMACS tools don't do.  You need to analyze water H-bonds with the two groups
of interest separately, then determine if the same water is H-bonded to a
moiety in both of those groups in the same frame.  This is where tracing the
H-bonds in the .xpm file is useful.

-Justin



Thanks for your comments, Justin.  When I invoke hbond by:

gmx hbond -f *.xtc -s *.tpr -num -hbn -g (± other options)

I get hbnum which shows changes with respect with time, but the hbond.ndx and
hbond.log files have no time information.  Are they averages of some sort?  Or a
particular frame that it defaults to? If so, how do I specific the frames/times?



Asking for -hbn requires -hbm to be useful.  The index file produced is the key 
for what the hydrogen bonds are in the .xpm file from -hbm.


-Justin

--
==

Justin A. Lemkul, Ph.D.
Ruth L. Kirschstein NRSA Postdoctoral Fellow

Department of Pharmaceutical Sciences
School of Pharmacy
Health Sciences Facility II, Room 629
University of Maryland, Baltimore
20 Penn St.
Baltimore, MD 21201

jalem...@outerbanks.umaryland.edu | (410) 706-7441
http://mackerell.umaryland.edu/~jalemkul

==
--
Gromacs Users mailing list

* Please search the archive at 
http://www.gromacs.org/Support/Mailing_Lists/GMX-Users_List before posting!

* Can't post? Read http://www.gromacs.org/Support/Mailing_Lists

* For (un)subscribe requests visit
https://maillist.sys.kth.se/mailman/listinfo/gromacs.org_gmx-users or send a 
mail to gmx-users-requ...@gromacs.org.

Re: [gmx-users] Computer processor and graphics card choice

[T.M.D. Graen]

80k atom MEM Benchmark System - 2fs time step - gmx2016.B2
GTX1070 - i7 6700k 4.00 GHz - 4 core - 42 ns/day - net cost per node (pc 
chassis) ~ 950 EUR
GTX1070 - Xeon E5 2630v4 2.20 - 10 core - 58 ns/day - net cost per node 
(19" rack) ~ 1600 EUR


GTX1080 will give you about the same performance but it is much more 
expensive, GTX1060 (6GB) is about 10% slower



On 12/01/2016 08:15 PM, Guillem Prats Ejarque wrote:

Dear colleagues,


I want to buy a computer to perform molecular dynamics by GROMACS, and I have 
several questions.


First of all, I am having doubts between the Intel Xeon E5-2620v4 (8 cores, 
2.1-3.0 GHz) and the i7-5820K (6 cores, 3.3-3.6 GHz). The reason of my doubts 
is the significantly lower clock rate of the Xeon processor. First I decided 
i7-5820k, but, although E5-2620 is at the limit of my budget, I would buy it 
only if the performance were significantly better.


Moreover, after the release of the new 10xx NVIDIA series, I also have doubts 
about the graphics card. First I wanted to buy the GTX 970 (1664 CUDA cores; 
1050-1178 MHz clock; 224 GB/s Memory bandwidth; 3494 SP Gflop/s), but looking 
at the new GTX 1060 - 6 GB (1280 CUDA cores; 1506-1708 MHz clock; 192 GB/s 
Memory bandwidth; 3855 SP Gflop/s) which is in the same price range,  it seemed 
to me that it could have higher performance with the last one. Would GTX 1060 
be better than GTX 970? Does anyone know if the 10xx series of NVIDA works well 
with GROMACS?


Thanks in advance,


Guillem



--
Gromacs Users mailing list

* Please search the archive at 
http://www.gromacs.org/Support/Mailing_Lists/GMX-Users_List before posting!

* Can't post? Read http://www.gromacs.org/Support/Mailing_Lists

* For (un)subscribe requests visit
https://maillist.sys.kth.se/mailman/listinfo/gromacs.org_gmx-users or send a 
mail to gmx-users-requ...@gromacs.org.

Re: [gmx-users] parallel processing

[Mark Abraham]

Hi,

Please check out
http://manual.gromacs.org/documentation/2016.1/user-guide/mdrun-performance.html.
Your first sample command is woefully inefficient, but for em this likely
doesn't matter.

Mark

On Fri, 2 Dec 2016 09:01 abhisek Mondal  wrote:

> But if I want to run the same job in 4 nodes (available cores= 4*16) then
> how would this work ?
>
> On Fri, Dec 2, 2016 at 2:20 PM,  wrote:
>
> > Hi Abhisek,
> >
> > You would need to use another version of gromacs with mpi rather than
> > thread-mpi (installed by adding DGMX_MPI=ON to the cmake command). You
> > could then use the following command
> >
> > mpirun -np 4 gmx_mpi mdrun -ntomp 16 -npme 0 -v -deffnm em
> >
> > I'm not sure why you are specifying -npme 0 but would suggest you don't
> do
> > this and instead let the number of separate pme ranks be set
> > automatically.
> >
> > Best wishes
> > James
> >
> > > Hi,
> > >
> > > I'm running gromacs on a cluster configuration as follows:
> > > 1 node = 16 cores
> > >
> > > I'm able to use single node with "gmx mdrun -ntmpi 4 -ntomp 16 -npme 0
> -v
> > > -deffnm em" command.
> > >
> > > How can I be able to run on multiple node (I have 20 nodes available) ?
> > > "-nt" is not doing good here.
> > >
> > >
> > >
> > > --
> > > Abhisek Mondal
> > >
> > > *Research Fellow*
> > >
> > > *Structural Biology and Bioinformatics Division*
> > > *CSIR-Indian Institute of Chemical Biology*
> > >
> > > *Kolkata 700032*
> > >
> > > *INDIA*
> > > --
> > > Gromacs Users mailing list
> > >
> > > * Please search the archive at
> > > http://www.gromacs.org/Support/Mailing_Lists/GMX-Users_List before
> > > posting!
> > >
> > > * Can't post? Read http://www.gromacs.org/Support/Mailing_Lists
> > >
> > > * For (un)subscribe requests visit
> > > https://maillist.sys.kth.se/mailman/listinfo/gromacs.org_gmx-users or
> > send
> > > a mail to gmx-users-requ...@gromacs.org.
> > >
> >
> >
> > --
> > Gromacs Users mailing list
> >
> > * Please search the archive at http://www.gromacs.org/
> > Support/Mailing_Lists/GMX-Users_List before posting!
> >
> > * Can't post? Read http://www.gromacs.org/Support/Mailing_Lists
> >
> > * For (un)subscribe requests visit
> > https://maillist.sys.kth.se/mailman/listinfo/gromacs.org_gmx-users or
> > send a mail to gmx-users-requ...@gromacs.org.
> >
>
>
>
> --
> Abhisek Mondal
>
> *Research Fellow*
>
> *Structural Biology and Bioinformatics Division*
> *CSIR-Indian Institute of Chemical Biology*
>
> *Kolkata 700032*
>
> *INDIA*
> --
> Gromacs Users mailing list
>
> * Please search the archive at
> http://www.gromacs.org/Support/Mailing_Lists/GMX-Users_List before
> posting!
>
> * Can't post? Read http://www.gromacs.org/Support/Mailing_Lists
>
> * For (un)subscribe requests visit
> https://maillist.sys.kth.se/mailman/listinfo/gromacs.org_gmx-users or
> send a mail to gmx-users-requ...@gromacs.org.
>
-- 
Gromacs Users mailing list

* Please search the archive at 
http://www.gromacs.org/Support/Mailing_Lists/GMX-Users_List before posting!

* Can't post? Read http://www.gromacs.org/Support/Mailing_Lists

* For (un)subscribe requests visit
https://maillist.sys.kth.se/mailman/listinfo/gromacs.org_gmx-users or send a 
mail to gmx-users-requ...@gromacs.org.

Re: [gmx-users] parallel processing

[abhisek Mondal]

But if I want to run the same job in 4 nodes (available cores= 4*16) then
how would this work ?

On Fri, Dec 2, 2016 at 2:20 PM,  wrote:

> Hi Abhisek,
>
> You would need to use another version of gromacs with mpi rather than
> thread-mpi (installed by adding DGMX_MPI=ON to the cmake command). You
> could then use the following command
>
> mpirun -np 4 gmx_mpi mdrun -ntomp 16 -npme 0 -v -deffnm em
>
> I'm not sure why you are specifying -npme 0 but would suggest you don't do
> this and instead let the number of separate pme ranks be set
> automatically.
>
> Best wishes
> James
>
> > Hi,
> >
> > I'm running gromacs on a cluster configuration as follows:
> > 1 node = 16 cores
> >
> > I'm able to use single node with "gmx mdrun -ntmpi 4 -ntomp 16 -npme 0 -v
> > -deffnm em" command.
> >
> > How can I be able to run on multiple node (I have 20 nodes available) ?
> > "-nt" is not doing good here.
> >
> >
> >
> > --
> > Abhisek Mondal
> >
> > *Research Fellow*
> >
> > *Structural Biology and Bioinformatics Division*
> > *CSIR-Indian Institute of Chemical Biology*
> >
> > *Kolkata 700032*
> >
> > *INDIA*
> > --
> > Gromacs Users mailing list
> >
> > * Please search the archive at
> > http://www.gromacs.org/Support/Mailing_Lists/GMX-Users_List before
> > posting!
> >
> > * Can't post? Read http://www.gromacs.org/Support/Mailing_Lists
> >
> > * For (un)subscribe requests visit
> > https://maillist.sys.kth.se/mailman/listinfo/gromacs.org_gmx-users or
> send
> > a mail to gmx-users-requ...@gromacs.org.
> >
>
>
> --
> Gromacs Users mailing list
>
> * Please search the archive at http://www.gromacs.org/
> Support/Mailing_Lists/GMX-Users_List before posting!
>
> * Can't post? Read http://www.gromacs.org/Support/Mailing_Lists
>
> * For (un)subscribe requests visit
> https://maillist.sys.kth.se/mailman/listinfo/gromacs.org_gmx-users or
> send a mail to gmx-users-requ...@gromacs.org.
>



-- 
Abhisek Mondal

*Research Fellow*

*Structural Biology and Bioinformatics Division*
*CSIR-Indian Institute of Chemical Biology*

*Kolkata 700032*

*INDIA*
-- 
Gromacs Users mailing list

* Please search the archive at 
http://www.gromacs.org/Support/Mailing_Lists/GMX-Users_List before posting!

* Can't post? Read http://www.gromacs.org/Support/Mailing_Lists

* For (un)subscribe requests visit
https://maillist.sys.kth.se/mailman/listinfo/gromacs.org_gmx-users or send a 
mail to gmx-users-requ...@gromacs.org.

Re: [gmx-users] parallel processing

[jkrieger]

Hi Abhisek,

You would need to use another version of gromacs with mpi rather than
thread-mpi (installed by adding DGMX_MPI=ON to the cmake command). You
could then use the following command

mpirun -np 4 gmx_mpi mdrun -ntomp 16 -npme 0 -v -deffnm em

I'm not sure why you are specifying -npme 0 but would suggest you don't do
this and instead let the number of separate pme ranks be set
automatically.

Best wishes
James

> Hi,
>
> I'm running gromacs on a cluster configuration as follows:
> 1 node = 16 cores
>
> I'm able to use single node with "gmx mdrun -ntmpi 4 -ntomp 16 -npme 0 -v
> -deffnm em" command.
>
> How can I be able to run on multiple node (I have 20 nodes available) ?
> "-nt" is not doing good here.
>
>
>
> --
> Abhisek Mondal
>
> *Research Fellow*
>
> *Structural Biology and Bioinformatics Division*
> *CSIR-Indian Institute of Chemical Biology*
>
> *Kolkata 700032*
>
> *INDIA*
> --
> Gromacs Users mailing list
>
> * Please search the archive at
> http://www.gromacs.org/Support/Mailing_Lists/GMX-Users_List before
> posting!
>
> * Can't post? Read http://www.gromacs.org/Support/Mailing_Lists
>
> * For (un)subscribe requests visit
> https://maillist.sys.kth.se/mailman/listinfo/gromacs.org_gmx-users or send
> a mail to gmx-users-requ...@gromacs.org.
>


-- 
Gromacs Users mailing list

* Please search the archive at 
http://www.gromacs.org/Support/Mailing_Lists/GMX-Users_List before posting!

* Can't post? Read http://www.gromacs.org/Support/Mailing_Lists

* For (un)subscribe requests visit
https://maillist.sys.kth.se/mailman/listinfo/gromacs.org_gmx-users or send a 
mail to gmx-users-requ...@gromacs.org.

Re: [gmx-users] Computer processor and graphics card choice

[Peter Kroon]

Hi Guille,


I'm not an expert on this, but it'll also depend on what kind of MD
simulations you want to run. Coarse-grained forcefields, for example,
don't benefit much from a GPU.

So what kind of simulations do you have in mind? Atomistic? QM/MM? How
many particles? ...?


Peter


On 02-12-16 00:56, Andrew Guy wrote:
> Hi Guillem,
>
> I can answer your last question - the 10xx series does indeed work well
> with GROMACS (currently running GROMACS 2016 with a GTX1070), although I
> would be inclined to get the GTX1070 if your budget can stretch to it.
>
> Andrew
>
> On Fri, Dec 2, 2016 at 6:15 AM, Guillem Prats Ejarque <
> guillem.prats.ejar...@uab.cat> wrote:
>
>> Dear colleagues,
>>
>>
>> I want to buy a computer to perform molecular dynamics by GROMACS, and I
>> have several questions.
>>
>>
>> First of all, I am having doubts between the Intel Xeon E5-2620v4 (8
>> cores, 2.1-3.0 GHz) and the i7-5820K (6 cores, 3.3-3.6 GHz). The reason of
>> my doubts is the significantly lower clock rate of the Xeon processor.
>> First I decided i7-5820k, but, although E5-2620 is at the limit of my
>> budget, I would buy it only if the performance were significantly better.
>>
>>
>> Moreover, after the release of the new 10xx NVIDIA series, I also have
>> doubts about the graphics card. First I wanted to buy the GTX 970 (1664
>> CUDA cores; 1050-1178 MHz clock; 224 GB/s Memory bandwidth; 3494 SP
>> Gflop/s), but looking at the new GTX 1060 - 6 GB (1280 CUDA cores;
>> 1506-1708 MHz clock; 192 GB/s Memory bandwidth; 3855 SP Gflop/s) which is
>> in the same price range,  it seemed to me that it could have higher
>> performance with the last one. Would GTX 1060 be better than GTX 970? Does
>> anyone know if the 10xx series of NVIDA works well with GROMACS?
>>
>>
>> Thanks in advance,
>>
>>
>> Guillem
>> --
>> Gromacs Users mailing list
>>
>> * Please search the archive at http://www.gromacs.org/
>> Support/Mailing_Lists/GMX-Users_List before posting!
>>
>> * Can't post? Read http://www.gromacs.org/Support/Mailing_Lists
>>
>> * For (un)subscribe requests visit
>> https://maillist.sys.kth.se/mailman/listinfo/gromacs.org_gmx-users or
>> send a mail to gmx-users-requ...@gromacs.org.
>>


-- 
Gromacs Users mailing list

* Please search the archive at 
http://www.gromacs.org/Support/Mailing_Lists/GMX-Users_List before posting!

* Can't post? Read http://www.gromacs.org/Support/Mailing_Lists

* For (un)subscribe requests visit
https://maillist.sys.kth.se/mailman/listinfo/gromacs.org_gmx-users or send a 
mail to gmx-users-requ...@gromacs.org.

Re: [gmx-users] Gomacs and multiple GPU

[Alexis Michon]

Hi,

Others software doesn't assist the user as gromacs, they don't use
CUDA_VISIBLE without an action by the user (with an option, or a
specific environnement variable), in case the use this variable they
wrote a bit more information, like what they detect and will be used. So
i was convince gromacs works in the same way even though you wrote in
your mail. sorry :(

Alexis

On 01/12/16 18:28, Szilárd Páll wrote:
> Hi,
>
> Welcome!
>
>> Thank for you reply. Yes, i have read them and know them. I use them
> will success with other software.
>
> BTW, what did you refer to here? It seems like you were focused on
> figuring out the detection issue and might have ignored my other
> recommendations on using multi (which incidentally would have solved
> the original issue too as well as would do correct job placement for
> you) and 2+ ranks per GPU?
> Cheers,
> --
> Szilárd
>
>
> On Thu, Dec 1, 2016 at 6:12 PM, Alexis Michon  wrote:
>> Hi,
>>
>> You solve my problem, Thank you.
>>
>> The variable CUDA_VISIBLE_DEVICES is set up by the scheduler at the
>> running time of my job. So after adding commandline "unset
>> $CUDA_VISIBLE_DEVICES" in my script, mdrun detect and use both GPUs.
>>
>> Thank you for your patience.
>>
>> Cheers,
>> Alexis
>>
>> On 01/12/16 17:53, Szilárd Páll wrote:
>>> Hi,
>>>
>>> Do any other programs, e.g. CUDA SDK samples (like deviceQuery) detect
>>> both GPUs?
>>> I find it *very-very* unlikely that only mdrun would not be able
>>> detect both GPUs.
>>>
>>> Again, are you sure it's not the use of CUDA_VISIBLE_DEVICES that
>>> creates the confusion? Please share your full command line (with env
>>> vars an all) and resulting log output?
>>>
>>> Cheers,
>>> --
>>> Szilárd
>>>
>>>
>>> On Thu, Dec 1, 2016 at 5:10 PM, Alexis Michon  wrote:
 Hi,

 Thank for you reply. Yes, i have read them and know them. I use them
 will success with other software.

 My first and main problem is mdrun doesn't see the second GPU on the
 system. And i try to find the root cause of and the remediation.

 Hardware is a bi xeon with 2 Titan-xp GPU,  so my goal is to run 2
 differentes simulations with two mdrun process each with one dedicated GPU.

 Do you have any idea ?

 Cheers
 Alexis

 On 01/12/16 16:38, Szilárd Páll wrote:
> Alexis,
>
> What is the problem? Have you read the relevant docs of what
> CUDA_VISIBLE_DEVICES does?
> https://docs.nvidia.com/cuda/cuda-c-programming-guide/#env-vars
>
>
> BTW,  you can use, but you *do not need* CUDA_VISIBLE devices to
> control the mapping of GPU(s) to mdrun process, the -gpu_id variable
> (or equivalent env var) alone is enough.
>
> To run multiple processes per node, do consider:
> - using mdrun -multi
> - using 2 or more mdrun runs per GPU
>
> You can find the reason, motivating performance examples, and sample
> commands here (in particular Fig 5 and related section):
> dx.doi.org/10.1002/jcc.24030
>
> Cheers,
> --
> Szilárd
>
>
> On Thu, Dec 1, 2016 at 3:38 PM, Alexis Michon  
> wrote:
>> Hi,
>>
>> thank you for your answer
>>
>> the Cmdline is:
>> gmx mdrun  -deffnm nvt
>>
>> I played a bit with option -gpu_id, use $CUDA_VISIBLE_DEVICES  and read
>> the page http://www.gromacs.org/GPU_acceleration without any succes. One
>> output is in nvt.log.11,
>>
>> nvidia-smi returs is in  nvidia-smi.log
>>
>> Nvidia-smi detect 2 GPUs, but not mdrun.
>>
>> nvidia-smi.log : https://icloud.ibcp.fr/index.php/s/08jqFsZpWNlSUGh
>> nvt.log : https://icloud.ibcp.fr/index.php/s/vC5UxCCAhyw9FuC
>>
>> Alexis
>>
>>
>>
>> On 01/12/16 11:34, Mark Abraham wrote:
>>> Hi,
>>>
>>> On Wed, Nov 30, 2016 at 5:38 PM Alexis Michon  
>>> wrote:
>>>
 Hello,

 We have build gromacs 2016.1 from source with "DMX_GPU=on" on a bi
 processor bi GPU machine, mdrun detect and run fine on only 1 GPU. How
 could we force mdrun to detect the second GPU ?

>>> If it's compatible, powered and supported by your driver, then mdrun 
>>> will
>>> find it. Presumably nvidia-smi tool will help you work out what's going 
>>> on.
>>>
>>> We would like to run 2 mdrun instances on a machine equiped with 2 GPUs,
 each mdrun will use 1GPU. How could we tell  mdrun to use a specific 
 GPU ?

>>> Guidance is here
>>> http://manual.gromacs.org/documentation/2016.1/user-guide/mdrun-performance.html#examples-for-mdrun-on-one-node
>>>
>>> Mark
>>>
>>>
 Cheers,
 Alexis

 --
 Citation : "It’s not enough to be busy; so are the ants. The question 
 is:
 what are we busy about?" - Henry