# PBC = xy slow performance

**URL:** <https://gromacs.bioexcel.eu/t/pbc-xy-slow-performance/8943>\
**Category:** User discussions\
**Created:** [April 21, 2024, 10:33am UTC](https://gromacs.bioexcel.eu/t/pbc-xy-slow-performance/8943 "2024-04-21T10:33:45Z")\
**Posts on this page:** 5\
**Page:** 1

<div class="post-metadata">

**Author:** ![GeorgePkhakadze](https://avatars.discourse-cdn.com/v4/letter/g/6a8cbe/32.png) [@GeorgePkhakadze](https://gromacs.bioexcel.eu/u/GeorgePkhakadze)\
**Post date:** [April 21, 2024, 10:33am UTC](https://gromacs.bioexcel.eu/t/pbc-xy-slow-performance/8943/1 "2024-04-21T10:33:45Z")

</div>

GROMACS version: 2024.1  
GROMACS modification: No  
Here post your question

Good time of the day. I’ve been running a membrane simulation using CGMartini in GROMACS and the simulation of PBC = xy is around 3 time slower than PBC = xyz with the same system.

Is there a way to improve performance?

---

<div class="post-metadata">

**Author:** ![GeorgePkhakadze](https://avatars.discourse-cdn.com/v4/letter/g/6a8cbe/32.png) [@GeorgePkhakadze](https://gromacs.bioexcel.eu/u/GeorgePkhakadze)\
**Post date:** [April 21, 2024, 10:35am UTC](https://gromacs.bioexcel.eu/t/pbc-xy-slow-performance/8943/2 "2024-04-21T10:35:00Z")

</div>

the .mdp parameters are:

title = Martini

integrator = md  
dt = 0.01  
nsteps = 34000000  
nstcomm = 100  
comm-grps =

nstxout = 0  
nstvout = 0  
nstfout = 0  
nstlog = 250000 ; Output frequency for energies to log file  
nstenergy = 250000 ; Output frequency for energies to energy file  
nstxtcout = 250000 ; Output frequency for .xtc file  
xtc\_precision = 100  
xtc-grps = POPC POPG W ION protein  
energygrps =

cutoff-scheme = Verlet  
nstlist = 150  
ns\_type = grid  
verlet-buffer-tolerance = 0.005

coulombtype = reaction-field  
rcoulomb = 1.1  
epsilon\_r = 15 ; 2.5 (with polarizable water)  
epsilon\_rf = 0  
vdw\_type = cutoff ;(for use with Verlet-pairlist)  
vdw-modifier = Potential-shift-verlet  
rvdw = 1.1 ;(for use with Verlet-pairlist)

tcoupl = v-rescale  
tc-grps = POPC POPG W ION protein  
tau\_t = 1.0 1.0 1.0 1.0 1.0  
ref\_t = 310 310 310 310 310

Pcoupl = no ; parrinello-rahman

gen\_vel = yes  
gen\_temp = 310  
gen\_seed = -1

constraints = none  
constraint\_algorithm = Lincs  
continuation = no  
lincs\_order = 4  
lincs\_warnangle = 30

pbc = xy  
nwall = 2  
wall-atomtype = W W  
wall-density = 6 6  
wall-r-linpot = -1

---

<div class="post-metadata">

**Author:** ![MagnusL](https://dub1.discourse-cdn.com/flex017/user_avatar/gromacs.bioexcel.eu/magnusl/32/2380_2.png) [@MagnusL](https://gromacs.bioexcel.eu/u/MagnusL)\
**Post date:** [April 22, 2024, 6:59am UTC](https://gromacs.bioexcel.eu/t/pbc-xy-slow-performance/8943/3 "2024-04-22T06:59:23Z")

</div>

Please don’t start multiple threads with the same questions ([PBC = xy slow performance](https://gromacs.bioexcel.eu/t/pbc-xy-slow-performance/8941)).

Have you checked the MD log files? Check what you are offloading to GPU. It’s possible that some interactions can still not be run on GPU with walls. Bonded interactions should be possible since 2024.1 ([Performance improvements - GROMACS 2024.1 documentation](https://manual.gromacs.org/current/release-notes/2024/major/performance.html#with-wall-potentials-bonded-interactions-can-now-be-run-on-gpus)), but there might still be other things that cannot be offloaded when using walls.

---

<div class="post-metadata">

**Author:** ![GeorgePkhakadze](https://avatars.discourse-cdn.com/v4/letter/g/6a8cbe/32.png) [@GeorgePkhakadze](https://gromacs.bioexcel.eu/u/GeorgePkhakadze)\
**Post date:** [April 22, 2024, 7:57am UTC](https://gromacs.bioexcel.eu/t/pbc-xy-slow-performance/8943/4 "2024-04-22T07:57:51Z")

</div>

Thank you for your reply. This is what I found in the log file:

Update task can not run on the GPU, because the following condition(s) were not satisfied:

Virtual sites are not supported.

From what I gather, the wall of the system is the virtual site that the log file mentions. According to the “real cycle and time accounting”, 75% of giga-cycles was spent on Force activity, which I’m guessing is the wall. Whereas with mega-flops accounting, “NxN RF Elec. + LJ [F]” accounted for 98% of flops.

---

<div class="post-metadata">

**Author:** ![MagnusL](https://dub1.discourse-cdn.com/flex017/user_avatar/gromacs.bioexcel.eu/magnusl/32/2380_2.png) [@MagnusL](https://gromacs.bioexcel.eu/u/MagnusL)\
**Post date:** [April 22, 2024, 9:00am UTC](https://gromacs.bioexcel.eu/t/pbc-xy-slow-performance/8943/5 "2024-04-22T09:00:32Z")

</div>

Virtual sites and walls are not the same.