# Does "nstvout" option make simulation crash?

**URL:** <https://gromacs.bioexcel.eu/t/does-nstvout-option-make-simulation-crash/6385>\
**Category:** User discussions\
**Tags:** mdp-parameters, mdrun\
**Created:** [May 11, 2023, 8:18pm UTC](https://gromacs.bioexcel.eu/t/does-nstvout-option-make-simulation-crash/6385 "2023-05-11T20:18:28Z")\
**Posts on this page:** 2\
**Page:** 1

<div class="post-metadata">

**Author:** ![congwang](https://avatars.discourse-cdn.com/v4/letter/c/9de0a6/32.png) [@congwang](https://gromacs.bioexcel.eu/u/congwang)\
**Post date:** [May 11, 2023, 8:18pm UTC](https://gromacs.bioexcel.eu/t/does-nstvout-option-make-simulation-crash/6385/1 "2023-05-11T20:18:28Z")

</div>

GROMACS version: 2022.2  
GROMACS modification: Yes/No  
Hi all,

My simulations encountered a really weird issue. They always crashed because of too many LINCS warnings. It can be solved by using a much smaller timestep. But the weird things is, when I set “nstvout” to any small vaules like 50000 but not the default value 0, all the LINCS warnings disappeared and the simulation can run smoothly with a regular timestemp. And if I set “nstvout” to a really large value like 1000000, the simulation crashed again.

In my understanding, “nstvout” is only about the frequency that we save velocities. I didn’t expect it can affect the simulation. Does anyone have an idea why?

(I’m doing a MARTINI simulation, you can find all the parameters, configurations and commands I used at [https://github.com/CongWang98/gromacs\_nstvout\_test](https://github.com/CongWang98/gromacs_nstvout_test))

---

<div class="post-metadata">

**Author:** ![congwang](https://avatars.discourse-cdn.com/v4/letter/c/9de0a6/32.png) [@congwang](https://gromacs.bioexcel.eu/u/congwang)\
**Post date:** [June 27, 2023, 6:41pm UTC](https://gromacs.bioexcel.eu/t/does-nstvout-option-make-simulation-crash/6385/2 "2023-06-27T18:41:22Z")

</div>

We used of different a different version of gromacs and the issue didnt happen anymore:)