[gmx-users] SIMD for gmx and sudden break in MDRUN

Justin Lemkul jalemkul at vt.edu
Sat Jul 13 19:52:22 CEST 2019



On 7/13/19 7:56 AM, kalpana wrote:
> Hi everyone,
>
> I installed gromacs with following configuration still my am getting
> suggestion mentioned bellow during mdrun. I used previous following command
> to restart. I used to use command (gmx mdrun md_0_1.tpr -deffnm md_0_1 -cpi
> md_0_1.cpt -append)  at sudden power breakdown. But it gives error. So
> given the situation, I have two questions:
> Q1. how to update configuration after installation as re-installation
> didn't help in this case.
> Q2. what is the difference between in both these two commands, I used one
> year back at the time of sudden break. Now I have forgotten and both are
> giving error.
>         gmx mdrun md_0_1.tpr -deffnm md_0_1 -cpi md_0_1.cpt -append
>         gmx mdrun deffnm -s md_0_1.tpr -cpi md_0_1.cpt -append

Neither of those commands is syntactically correct.

gmx mdrun -deffnm md_0_1 -cpi -append

>
> *Cmake configuration*
> sudo cmake .. -DGMX_BUILD_OWN_FFTW=OFF -DREGRESSIONTEST_DOWNLOAD=OFF
> -DCMAKE_C_COMPILER=gcc -DGMX_MPI=on -DGMX_USE_RDTSCP=ON -DGMX_GPU=on
> -DGMX_SIMD=AVX2_256 -DGMX_FFT_LIBRARY=fftw3
> -DREGRESSIONTEST_PATH=/gromacs/regressiontests-2019.3
>
> *Suggestion during MDRUN*
> Compiled SIMD: SSE2, but for this host/run AVX2_256 might be better (see
> log).

Here, mdrun is telling you that it is slower than it could be on the 
given hardware. You should let cmake automatically find the best SIMD 
optimization.

> The current CPU can measure timings more accurately than the code in
> gmx mdrun was configured to use. This might affect your simulation
> speed as accurate timings are needed for load-balancing.
> Please consider rebuilding gmx mdrun with the GMX_USE_RDTSCP=ON CMake
> option.

Again, use this setting when compiling.

Neither of these messages will halt mdrun. They are merely advising you 
that you are not getting the best possible performance.

-Justin

> Reading file em.tpr, VERSION 2019.3 (single precision)
>
> Thanks & best regards,
> Kalpana

-- 
==================================================

Justin A. Lemkul, Ph.D.
Assistant Professor
Office: 301 Fralin Hall
Lab: 303 Engel Hall

Virginia Tech Department of Biochemistry
340 West Campus Dr.
Blacksburg, VA 24061

jalemkul at vt.edu | (540) 231-3129
http://www.thelemkullab.com

==================================================



More information about the gromacs.org_gmx-users mailing list