# Not able to see any acceleration when using two nodes **URL:** https://gromacs.bioexcel.eu/t/not-able-to-see-any-acceleration-when-using-two-nodes/1951 **Category:** User discussions **Tags:** mdrun **Created:** [April 13, 2021, 1:44am UTC](https://gromacs.bioexcel.eu/t/not-able-to-see-any-acceleration-when-using-two-nodes/1951 "2021-04-13T01:44:34Z") **Posts on this page:** 1 **Page:** 1
### Author: ![qingsong](https://avatars.discourse-cdn.com/v4/letter/q/54ee81/32.png) [@qingsong](https://gromacs.bioexcel.eu/u/qingsong) #### Post date: [April 13, 2021, 1:44am UTC](https://gromacs.bioexcel.eu/t/not-able-to-see-any-acceleration-when-using-two-nodes/1951/1 "2021-04-13T01:44:34Z")
GROMACS version: 2019. 06 I’m using GROMACS 2019.06 version with intel-mpi enabled to test its performance, the weird thing is it is slower when using two nodes comparing to using 1 node to run the same simulation, here comes configuration: Node Configuration: 96 cpu, 4 T4 gpu for each node Node Names: g-t4-4-worker0001 and g-t4-4-worker0002. Run Configuration: 8 MPI processes, 24 openmp threads per MPI process. SLURM log: [0] MPI startup(): Multi-threaded optimized library [0] DAPL startup: RLIMIT\_MEMLOCK too small [1] DAPL startup: RLIMIT\_MEMLOCK too small [2] DAPL startup: RLIMIT\_MEMLOCK too small [3] DAPL startup: RLIMIT\_MEMLOCK too small [4] DAPL startup: RLIMIT\_MEMLOCK too small [5] DAPL startup: RLIMIT\_MEMLOCK too small [6] DAPL startup: RLIMIT\_MEMLOCK too small [7] DAPL startup: RLIMIT\_MEMLOCK too small [4] MPI startup(): cannot load default tmi provider [5] MPI startup(): cannot load default tmi provider [6] MPI startup(): cannot load default tmi provider [7] MPI startup(): cannot load default tmi provider [0] MPI startup(): cannot load default tmi provider [1] MPI startup(): cannot load default tmi provider [2] MPI startup(): cannot load default tmi provider [3] MPI startup(): cannot load default tmi provider [4] MPI startup(): RLIMIT\_MEMLOCK too small [5] MPI startup(): RLIMIT\_MEMLOCK too small [6] MPI startup(): RLIMIT\_MEMLOCK too small [7] MPI startup(): RLIMIT\_MEMLOCK too small [0] MPI startup(): RLIMIT\_MEMLOCK too small [1] MPI startup(): RLIMIT\_MEMLOCK too small [2] MPI startup(): RLIMIT\_MEMLOCK too small [3] MPI startup(): RLIMIT\_MEMLOCK too small [2] MPI startup(): shm and tcp data transfer modes [0] MPI startup(): shm and tcp data transfer modes [3] MPI startup(): shm and tcp data transfer modes [1] MPI startup(): shm and tcp data transfer modes [5] MPI startup(): shm and tcp data transfer modes [6] MPI startup(): shm and tcp data transfer modes [4] MPI startup(): shm and tcp data transfer modes [7] MPI startup(): shm and tcp data transfer modes [0] MPI startup(): Rank Pid Node name Pin cpu [0] MPI startup(): 0 15026 g-t4-4-worker0001 {0,1,2,3,4,5,6,7,8,9,10,11,12,13,14,15,16,17,18,19,20,21,22,23} [0] MPI startup(): 1 15027 g-t4-4-worker0001 {24,25,26,27,28,29,30,31,32,33,34,35,36,37,38,39,40,41,42,43,44,45,46,47} [0] MPI startup(): 2 15028 g-t4-4-worker0001 {48,49,50,51,52,53,54,55,56,57,58,59,60,61,62,63,64,65,66,67,68,69,70,71} [0] MPI startup(): 3 15029 g-t4-4-worker0001 {72,73,74,75,76,77,78,79,80,81,82,83,84,85,86,87,88,89,90,91,92,93,94,95} [0] MPI startup(): 4 15042 g-t4-4-worker0002 {0,1,2,3,4,5,6,7,8,9,10,11,12,13,14,15,16,17,18,19,20,21,22,23} [0] MPI startup(): 5 15043 g-t4-4-worker0002 {24,25,26,27,28,29,30,31,32,33,34,35,36,37,38,39,40,41,42,43,44,45,46,47} [0] MPI startup(): 6 15044 g-t4-4-worker0002 {48,49,50,51,52,53,54,55,56,57,58,59,60,61,62,63,64,65,66,67,68,69,70,71} [0] MPI startup(): 7 15045 g-t4-4-worker0002 {72,73,74,75,76,77,78,79,80,81,82,83,84,85,86,87,88,89,90,91,92,93,94,95} [0] MPI startup(): I\_MPI\_DEBUG=5 [0] MPI startup(): I\_MPI\_HYDRA\_UUID=ac3a0000-aeee-ff9b-b0bf-0500e05d0a08 [0] MPI startup(): I\_MPI\_INFO\_NUMA\_NODE\_NUM=1 [0] MPI startup(): I\_MPI\_PIN\_MAPPING=4:0 0,1 24,2 48,3 72 :-) GROMACS - gmx mdrun, 2019.6 (-: ``` GROMACS is written by: Emile Apol Rossen Apostolov Paul Bauer Herman J.C. Berendsen Par Bjelkmar Christian Blau Viacheslav Bolnykh Kevin Boyd ``` Aldert van Buuren Rudi van Drunen Anton Feenstra Alan Gray Gerrit Groenhof Anca Hamuraru Vincent Hindriksen M. Eric Irrgang Aleksei Iupinov Christoph Junghans Joe Jordan Dimitrios Karkoulis Peter Kasson Jiri Kraus Carsten Kutzner Per Larsson Justin A. Lemkul Viveca Lindahl Magnus Lundborg Erik Marklund Pascal Merz Pieter Meulenhoff Teemu Murtola Szilard Pall Sander Pronk Roland Schulz Michael Shirts Alexey Shvetsov Alfons Sijbers Peter Tieleman Jon Vincent Teemu Virolainen Christian Wennberg Maarten 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) 2001-2018, The GROMACS development team at Uppsala University, Stockholm University and the Royal Institute of Technology, Sweden. check out [http://www.gromacs.org](http://www.gromacs.org) for more information. GROMACS is free software; you can redistribute it and/or modify it under the terms of the GNU Lesser General Public License as published by the Free Software Foundation; either version 2.1 of the License, or (at your option) any later version. GROMACS: gmx mdrun, version 2019.6 Executable: /public/software/.local/easybuild/software/GROMACS/2019.6-intelmpi-cuda/bin/gmx\_mpi Data prefix: /public/software/.local/easybuild/software/GROMACS/2019.6-intelmpi-cuda Working dir: /home/cloudam/gromacs-demo/gromacs\_test/MD Command line: gmx\_mpi mdrun -v -pin on -nb gpu -bonded gpu -pme gpu -npme 1 -ntomp 24 -deffnm md\_100 Back Off! I just backed up md\_100.log to ./#md\_100.log.53# Compiled SIMD: AVX2\_256, but for this host/run AVX\_512 might be better (see log). Reading file md\_100.tpr, VERSION 2019.6 (single precision) Changing nstlist from 20 to 100, rlist from 1.228 to 1.356 Using 8 MPI processes Using 24 OpenMP threads per MPI process On host g-t4-4-worker0001 4 GPUs selected for this run. Mapping of GPU IDs to the 4 GPU tasks in the 4 ranks on this node: PP:0,PP:1,PP:2,PP:3 PP tasks will do (non-perturbed) short-ranged and most bonded interactions on the GPU PME tasks will do all aspects on the GPU NOTE: Your choice of number of MPI ranks and amount of resources results in using 24 OpenMP threads per rank, which is most likely inefficient. The optimum is usually between 2 and 6 threads per rank. Overriding thread affinity set outside gmx mdrun Back Off! I just backed up md\_100.xtc to ./#md\_100.xtc.41# Back Off! I just backed up md\_100.edr to ./#md\_100.edr.41# WARNING: This run will generate roughly 29602 Mb of data NOTE: DLB will not turn on during the first phase of PME tuning starting mdrun ‘Protein in water’ 50000000 steps, 100000.0 ps. step 0 imb F 11% pme/F 1.02 step 100, will finish Sat Apr 17 13:40:41 2021 imb F 12% pme/F 0.94 step 200, will finish Sat Apr 17 05:34:43 2021 imb F 12% pme/F 0.96 step 300, will finish Sat Apr 17 02:42:06 2021 imb F 11% pme/F 0.94 step 400, will finish Sat Apr 17 01:28:01 2021 imb F 12% pme/F 0.96 step 500, will finish Sat Apr 17 00:32:33 2021 imb F 12% pme/F 0.95 step 600, will finish Sat Apr 17 00:08:24 2021 imb F 16% pme/F 0.89 step 700, will finish Fri Apr 16 23:41:09 2021 imb F 9% pme/F 0.99 step 800, will finish Fri Apr 16 23:30:44 2021 imb F 12% pme/F 0.98 step 900, will finish Fri Apr 16 23:21:05 2021 imb F 12% pme/F 0.95 step 1000, will finish Fri Apr 16 23:44:59 2021 imb F 10% pme/F 0.96 step 1100, will finish Fri Apr 16 23:45:08 2021 imb F 13% pme/F 0.95 step 1200, will finish Fri Apr 16 23:36:00 2021 imb F 13% pme/F 0.96 step 1300, will finish Fri Apr 16 23:26:45 2021 imb F 13% pme/F 0.95 step 1400, will finish Fri Apr 16 23:15:02 2021 imb F 13% pme/F 0.92 step 1500, will finish Fri Apr 16 23:09:06 2021 imb F 17% pme/F 0.88 step 1600, will finish Fri Apr 16 23:00:16 2021 imb F 16% pme/F 0.89 step 1700, will finish Fri Apr 16 22:55:14 2021 imb F 11% pme/F 0.95 step 1800, will finish Fri Apr 16 22:51:26 2021 imb F 10% pme/F 0.98 step 1900, will finish Fri Apr 16 22:45:44 2021 imb F 14% pme/F 0.95 step 2000, will finish Fri Apr 16 22:41:10 2021 imb F 18% pme/F 0.91 step 2100, will finish Fri Apr 16 22:33:45 2021 imb F 12% pme/F 0.93 step 2200, will finish Fri Apr 16 22:24:03 2021 imb F 12% pme/F 0.94 step 2300, will finish Fri Apr 16 22:17:08 2021 imb F 11% pme/F 0.94 step 2400, will finish Fri Apr 16 22:12:50 2021 imb F 16% pme/F 0.90 step 2500, will finish Fri Apr 16 22:23:02 2021 imb F 11% pme/F 0.97 step 2600, will finish Fri Apr 16 22:17:49 2021 imb F 12% pme/F 0.94 step 2700, will finish Fri Apr 16 22:16:54 2021 imb F 14% pme/F 0.91 step 2800, will finish Fri Apr 16 22:17:57 2021 imb F 10% pme/F 0.96 step 2900, will finish Fri Apr 16 22:13:35 2021 imb F 12% pme/F 0.95 step 3000, will finish Fri Apr 16 22:11:23 2021 imb F 12% pme/F 0.93 step 3100, will finish Fri Apr 16 22:11:33 2021 imb F 13% pme/F 0.97 step 3200, will finish Fri Apr 16 22:08:39 2021 imb F 10% pme/F 0.99 step 3300, will finish Fri Apr 16 22:08:44 2021 imb F 12% pme/F 0.98 step 3400, will finish Fri Apr 16 22:17:20 2021 imb F 11% pme/F 0.98 step 3500, will finish Fri Apr 16 22:16:02 2021 imb F 18% pme/F 0.91 step 3600, will finish Fri Apr 16 22:12:21 2021 imb F 18% pme/F 0.88 step 3700, will finish Fri Apr 16 22:10:23 2021 imb F 12% pme/F 0.94 step 3800, will finish Fri Apr 16 22:08:45 2021 imb F 17% pme/F 0.90 step 3900, will finish Fri Apr 16 22:08:25 2021 MD.log: GROMACS: gmx mdrun, version 2019.6 Executable: /public/software/.local/easybuild/software/GROMACS/2019.6-intelmpi-cuda/bin/gmx\_mpi Data prefix: /public/software/.local/easybuild/software/GROMACS/2019.6-intelmpi-cuda Working dir: /home/cloudam/gromacs-demo/gromacs\_test/MD Process ID: 15026 Command line: gmx\_mpi mdrun -v -pin on -nb gpu -bonded gpu -pme gpu -npme 1 -ntomp 24 -deffnm md\_100 GROMACS version: 2019.6 Precision: single Memory model: 64 bit MPI library: MPI OpenMP support: enabled (GMX\_OPENMP\_MAX\_THREADS = 64) GPU support: CUDA SIMD instructions: AVX2\_256 FFT library: fftw-3.3.8-sse2-avx-avx2-avx2\_128 RDTSCP usage: enabled TNG support: enabled Hwloc support: hwloc-1.11.12 Tracing support: disabled C compiler: /public/software/.local/easybuild/software/impi/2018.5.288-iccifort-2019.5.281/bin64/mpicc GNU 8.3.0 C compiler flags: -mavx2 -mfma -O3 -DNDEBUG -funroll-all-loops -fexcess-precision=fast C++ compiler: /public/software/.local/easybuild/software/impi/2018.5.288-iccifort-2019.5.281/bin64/mpicxx GNU 8.3.0 C++ compiler flags: -mavx2 -mfma -std=c++11 -O3 -DNDEBUG -funroll-all-loops -fexcess-precision=fast CUDA compiler: /public/software/.local/easybuild/software/CUDA/10.1.243-iccifort-2019.5.281/bin/nvcc nvcc: NVIDIA (R) Cuda compiler driver;Copyright (c) 2005-2019 NVIDIA Corporation;Built on Sun\_Jul\_28\_19:07:16\_PDT\_2019;Cuda compilation tools, release 10.1, V10.1.243 CUDA compiler flags:-gencode;arch=compute\_30,code=sm\_30;-gencode;arch=compute\_35,code=sm\_35;-gencode;arch=compute\_37,code=sm\_37;-gencode;arch=compute\_50,code=sm\_50;-gencode;arch=compute\_52,code=sm\_52;-gencode;arch=compute\_60,code=sm\_60;-gencode;arch=compute\_61,code=sm\_61;-gencode;arch=compute\_70,code=sm\_70;-gencode;arch=compute\_75,code=compute\_75;-use\_fast\_math;;; ;-mavx2;-mfma;-std=c++11;-O3;-DNDEBUG;-funroll-all-loops;-fexcess-precision=fast; CUDA driver: 10.10 CUDA runtime: 10.10 Running on 2 nodes with total 96 cores, 192 logical cores, 8 compatible GPUs Cores per node: 48 Logical cores per node: 96 Compatible GPUs per node: 4 All nodes have identical type(s) of GPUs Hardware detected on host g-t4-4-worker0001 (the node of MPI rank 0): CPU info: Vendor: Intel Brand: Intel(R) Xeon(R) Platinum 8163 CPU @ 2.50GHz Family: 6 Model: 85 Stepping: 4 Features: aes apic avx avx2 avx512f avx512cd avx512bw avx512vl clfsh cmov cx8 cx16 f16c fma hle htt intel lahf mmx msr nonstop\_tsc pcid pclmuldq pdpe1gb popcnt pse rdrnd rdtscp rtm sse2 sse3 sse4.1 sse4.2 ssse3 x2apic Number of AVX-512 FMA units: 2 Hardware topology: Full, with devices Sockets, cores, and logical processors: Socket 0: [0 1] [2 3] [4 5] [6 7] [8 9] [10 11] [12 13] [14 15] [16 17] [18 19] [20 21] [22 23] [24 25] [26 27] [28 29] [30 31] [32 33] [34 35] [36 37] [38 39] [40 41] [42 43] [44 45] [46 47] [48 49] [50 51] [52 53] [54 55] [56 57] [58 59] [60 61] [62 63] [64 65] [66 67] [68 69] [70 71] [72 73] [74 75] [76 77] [78 79] [80 81] [82 83] [84 85] [86 87] [88 89] [90 91] [92 93] [94 95] Numa nodes: Node 0 (392897486848 bytes mem): 0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 Latency: 0 0 1.00 Caches: L1: 32768 bytes, linesize 64 bytes, assoc. 8, shared 2 ways L2: 1048576 bytes, linesize 64 bytes, assoc. 16, shared 2 ways L3: 34603008 bytes, linesize 64 bytes, assoc. 11, shared 64 ways PCI devices: 0000:00:01.1 Id: 8086:7010 Class: 0x0101 Numa: 0 0000:00:02.0 Id: 1013:00b8 Class: 0x0300 Numa: 0 0000:00:05.0 Id: 1af4:1001 Class: 0x0100 Numa: 0 0000:00:06.0 Id: 1af4:1000 Class: 0x0200 Numa: 0 0000:00:07.0 Id: 10de:1eb8 Class: 0x0302 Numa: 0 0000:00:08.0 Id: 10de:1eb8 Class: 0x0302 Numa: 0 0000:00:09.0 Id: 10de:1eb8 Class: 0x0302 Numa: 0 0000:00:0a.0 Id: 10de:1eb8 Class: 0x0302 Numa: 0 GPU info: Number of GPUs detected: 4 #0: NVIDIA Tesla T4, compute cap.: 7.5, ECC: yes, stat: compatible #1: NVIDIA Tesla T4, compute cap.: 7.5, ECC: yes, stat: compatible #2: NVIDIA Tesla T4, compute cap.: 7.5, ECC: yes, stat: compatible #3: NVIDIA Tesla T4, compute cap.: 7.5, ECC: yes, stat: compatible Highest SIMD level requested by all nodes in run: AVX\_512 SIMD instructions selected at compile time: AVX2\_256 This program was compiled for different hardware than you are running on, which could influence performance. This build might have been configured on a login node with only a single AVX-512 FMA unit (in which case AVX2 is faster), while the node you are running on has dual AVX-512 FMA units. The number of OpenMP threads was set by environment variable OMP\_NUM\_THREADS to 24 (and the command-line setting agreed with that) Input Parameters: integrator = md tinit = 0 dt = 0.002 nsteps = 50000000 init-step = 0 simulation-part = 1 comm-mode = Linear nstcomm = 100 bd-fric = 0 ld-seed = -1222564944 emtol = 10 emstep = 0.01 niter = 20 fcstep = 0 nstcgsteep = 1000 nbfgscorr = 10 rtpi = 0.05 nstxout = 0 nstvout = 0 nstfout = 0 nstlog = 2500 nstcalcenergy = 100 nstenergy = 2500 nstxout-compressed = 2500 compressed-x-precision = 1000 cutoff-scheme = Verlet nstlist = 20 ns-type = Grid pbc = xyz periodic-molecules = false verlet-buffer-tolerance = 0.005 rlist = 1.228 coulombtype = PME coulomb-modifier = Potential-shift rcoulomb-switch = 0 rcoulomb = 1.2 epsilon-r = 1 epsilon-rf = inf vdw-type = Cut-off vdw-modifier = Force-switch rvdw-switch = 1 rvdw = 1.2 DispCorr = EnerPres table-extension = 1 fourierspacing = 0.16 fourier-nx = 96 fourier-ny = 96 fourier-nz = 96 pme-order = 4 ewald-rtol = 1e-05 ewald-rtol-lj = 0.001 lj-pme-comb-rule = Geometric ewald-geometry = 0 epsilon-surface = 0 tcoupl = V-rescale nsttcouple = 20 nh-chain-length = 0 print-nose-hoover-chain-variables = false pcoupl = Parrinello-Rahman pcoupltype = Isotropic nstpcouple = 20 tau-p = 2 compressibility (3x3): compressibility[0]={ 4.50000e-05, 0.00000e+00, 0.00000e+00} compressibility[1]={ 0.00000e+00, 4.50000e-05, 0.00000e+00} compressibility[2]={ 0.00000e+00, 0.00000e+00, 4.50000e-05} ref-p (3x3): ref-p[0]={ 1.00000e+00, 0.00000e+00, 0.00000e+00} ref-p[1]={ 0.00000e+00, 1.00000e+00, 0.00000e+00} ref-p[2]={ 0.00000e+00, 0.00000e+00, 1.00000e+00} refcoord-scaling = No posres-com (3): posres-com[0]= 0.00000e+00 posres-com[1]= 0.00000e+00 posres-com[2]= 0.00000e+00 posres-comB (3): posres-comB[0]= 0.00000e+00 posres-comB[1]= 0.00000e+00 posres-comB[2]= 0.00000e+00 QMMM = false QMconstraints = 0 QMMMscheme = 0 MMChargeScaleFactor = 1 qm-opts: ngQM = 0 constraint-algorithm = Lincs continuation = true Shake-SOR = false shake-tol = 0.0001 lincs-order = 4 lincs-iter = 1 lincs-warnangle = 30 nwall = 0 wall-type = 9-3 wall-r-linpot = -1 wall-atomtype[0] = -1 wall-atomtype[1] = -1 wall-density[0] = 0 wall-density[1] = 0 wall-ewald-zfac = 3 pull = false awh = false rotation = false interactiveMD = false disre = No disre-weighting = Conservative disre-mixed = false dr-fc = 1000 dr-tau = 0 nstdisreout = 100 orire-fc = 0 orire-tau = 0 nstorireout = 100 free-energy = no cos-acceleration = 0 deform (3x3): deform[0]={ 0.00000e+00, 0.00000e+00, 0.00000e+00} deform[1]={ 0.00000e+00, 0.00000e+00, 0.00000e+00} deform[2]={ 0.00000e+00, 0.00000e+00, 0.00000e+00} simulated-tempering = false swapcoords = no userint1 = 0 userint2 = 0 userint3 = 0 userint4 = 0 userreal1 = 0 userreal2 = 0 userreal3 = 0 userreal4 = 0 applied-forces: electric-field: x: E0 = 0 omega = 0 t0 = 0 sigma = 0 y: E0 = 0 omega = 0 t0 = 0 sigma = 0 z: E0 = 0 omega = 0 t0 = 0 sigma = 0 grpopts: nrdf: 28794.8 447873 ref-t: 300 300 tau-t: 0.1 0.1 annealing: No No annealing-npoints: 0 0 acc: 0 0 0 nfreeze: N N N energygrp-flags[0]: 0 Changing nstlist from 20 to 100, rlist from 1.228 to 1.356 Initializing Domain Decomposition on 8 ranks Dynamic load balancing: locked Minimum cell size due to atom displacement: 0.762 nm Initial maximum distances in bonded interactions: two-body bonded interactions: 0.438 nm, LJ-14, atoms 8608 8616 multi-body bonded interactions: 0.438 nm, Ryckaert-Bell., atoms 8608 8616 Minimum cell size due to bonded interactions: 0.482 nm Maximum distance for 5 constraints, at 120 deg. angles, all-trans: 0.218 nm Estimated maximum distance required for P-LINCS: 0.218 nm Using 1 separate PME ranks Scaling the initial minimum size with 1/0.8 (option -dds) = 1.25 Optimizing the DD grid for 7 cells with a minimum initial size of 0.953 nm The maximum allowed number of cells is: X 12 Y 12 Z 11 Domain decomposition grid 7 x 1 x 1, separate PME ranks 1 PME domain decomposition: 1 x 1 x 1 Interleaving PP and PME ranks This rank does only particle-particle work. Domain decomposition rank 0, coordinates 0 0 0 The initial number of communication pulses is: X 1 The initial domain decomposition cell size is: X 1.74 nm The maximum allowed distance for atoms involved in interactions is: non-bonded interactions 1.356 nm (the following are initial values, they could change due to box deformation) two-body bonded interactions (-rdd) 1.356 nm multi-body bonded interactions (-rdd) 1.356 nm virtual site constructions (-rcon) 1.740 nm atoms separated by up to 5 constraints (-rcon) 1.740 nm When dynamic load balancing gets turned on, these settings will change to: The maximum number of communication pulses is: X 1 The minimum size for domain decomposition cells is 1.356 nm The requested allowed shrink of DD cells (option -dds) is: 0.80 The allowed shrink of domain decomposition cells is: X 0.78 The maximum allowed distance for atoms involved in interactions is: non-bonded interactions 1.356 nm two-body bonded interactions (-rdd) 1.356 nm multi-body bonded interactions (-rdd) 1.356 nm virtual site constructions (-rcon) 1.356 nm atoms separated by up to 5 constraints (-rcon) 1.356 nm Using two step summing over 2 groups of on average 3.5 ranks Using 8 MPI processes Using 24 OpenMP threads per MPI process On host g-t4-4-worker0001 4 GPUs selected for this run. Mapping of GPU IDs to the 4 GPU tasks in the 4 ranks on this node: PP:0,PP:1,PP:2,PP:3 PP tasks will do (non-perturbed) short-ranged and most bonded interactions on the GPU PME tasks will do all aspects on the GPU NOTE: Your choice of number of MPI ranks and amount of resources results in using 24 OpenMP threads per rank, which is most likely inefficient. The optimum is usually between 2 and 6 threads per rank. Overriding thread affinity set outside gmx mdrun Pinning threads with an auto-selected logical core stride of 1 System total charge: 0.000 Will do PME sum in reciprocal space for electrostatic interactions. ++++ PLEASE READ AND CITE THE FOLLOWING REFERENCE ++++ U. Essmann, L. Perera, M. L. Berkowitz, T. Darden, H. Lee and L. G. Pedersen A smooth particle mesh Ewald method J. Chem. Phys. 103 (1995) pp. 8577-8592 -------- -------- — Thank You — -------- -------- Using a Gaussian width (1/beta) of 0.384195 nm for Ewald Potential shift: LJ r^-12: -2.648e-01 r^-6: -5.349e-01, Ewald -8.333e-06 Initialized non-bonded Ewald correction tables, spacing: 1.02e-03 size: 1176 Long Range LJ corr.: 1.7854e-04 Generated table with 1178 data points for Ewald. Tabscale = 500 points/nm Generated table with 1178 data points for LJ6Shift. Tabscale = 500 points/nm Generated table with 1178 data points for LJ12Shift. Tabscale = 500 points/nm Generated table with 1178 data points for 1-4 COUL. Tabscale = 500 points/nm Generated table with 1178 data points for 1-4 LJ6. Tabscale = 500 points/nm Generated table with 1178 data points for 1-4 LJ12. Tabscale = 500 points/nm Using GPU 8x8 nonbonded short-range kernels Using a dual 8x4 pair-list setup updated with dynamic, rolling pruning: outer list: updated every 100 steps, buffer 0.156 nm, rlist 1.356 nm inner list: updated every 10 steps, buffer 0.002 nm, rlist 1.202 nm At tolerance 0.005 kJ/mol/ps per atom, equivalent classical 1x1 list would be: outer list: updated every 100 steps, buffer 0.307 nm, rlist 1.507 nm inner list: updated every 10 steps, buffer 0.044 nm, rlist 1.244 nm Initializing Parallel LINear Constraint Solver ++++ PLEASE READ AND CITE THE FOLLOWING REFERENCE ++++ B. Hess P-LINCS: A Parallel Linear Constraint Solver for molecular simulation J. Chem. Theory Comput. 4 (2008) pp. 116-122 -------- -------- — Thank You — -------- -------- The number of constraints is 5765 There are constraints between atoms in different decomposition domains, will communicate selected coordinates each lincs iteration ++++ PLEASE READ AND CITE THE FOLLOWING REFERENCE ++++ S. Miyamoto and P. A. Kollman SETTLE: An Analytical Version of the SHAKE and RATTLE Algorithms for Rigid Water Models J. Comp. Chem. 13 (1992) pp. 952-962 -------- -------- — Thank You — -------- -------- Linking all bonded interactions to atoms There are 74428 inter charge-group virtual sites, will an extra communication step for selected coordinates and forces Intra-simulation communication will occur every 20 steps. Center of mass motion removal mode is Linear We have the following groups for center of mass motion removal: 0: rest ++++ PLEASE READ AND CITE THE FOLLOWING REFERENCE ++++ G. Bussi, D. Donadio and M. Parrinello Canonical sampling through velocity rescaling J. Chem. Phys. 126 (2007) pp. 014101 -------- -------- — Thank You — -------- -------- There are: 235240 Atoms There are: 74428 VSites Atom distribution over 7 domains: av 44238 stddev 277 min 43920 max 44619 NOTE: DLB will not turn on during the first phase of PME tuning Started mdrun on rank 0 Sun Apr 11 19:22:27 2021 ``` Step Time 0 0.00000 ``` Energies (kJ/mol) Bond Harmonic Pot. Angle Ryckaert-Bell. Improper Dih. 8.41905e+03 9.05915e+00 2.29383e+04 1.37466e+04 1.56581e+03 LJ-14 Coulomb-14 LJ (SR) Disper. corr. Coulomb (SR) 1.46827e+04 6.84097e+04 5.48357e+05 -2.21763e+04 -4.04786e+06 Coul. recip. Potential Kinetic En. Total Energy Conserved En. 1.16819e+04 -3.38023e+06 5.92384e+05 -2.78784e+06 -2.78770e+06 Temperature Pres. DC (bar) Pressure (bar) Constr. rmsd 2.98939e+02 -1.57011e+02 3.95213e+01 4.56933e-06 DD step 99 load imb.: force 11.2% pme mesh/force 1.020 DD step 2499 load imb.: force 16.2% pme mesh/force 0.895 Step Time 2500 5.00000 Energies (kJ/mol) Bond Harmonic Pot. Angle Ryckaert-Bell. Improper Dih. 8.47799e+03 1.78844e+01 2.30134e+04 1.37323e+04 1.65879e+03 LJ-14 Coulomb-14 LJ (SR) Disper. corr. Coulomb (SR) 1.48272e+04 6.85086e+04 5.52541e+05 -2.22215e+04 -4.05390e+06 Coul. recip. Potential Kinetic En. Total Energy Conserved En. 1.14112e+04 -3.38194e+06 5.93595e+05 -2.78834e+06 -2.78767e+06 Temperature Pres. DC (bar) Pressure (bar) Constr. rmsd 2.99551e+02 -1.57651e+02 1.18464e+02 4.44390e-06 DD step 4999 load imb.: force 15.1% pme mesh/force 0.952 Step Time 5000 10.00000 Energies (kJ/mol) Bond Harmonic Pot. Angle Ryckaert-Bell. Improper Dih. 8.32716e+03 1.96349e+01 2.33778e+04 1.35488e+04 1.50029e+03 LJ-14 Coulomb-14 LJ (SR) Disper. corr. Coulomb (SR) 1.46913e+04 6.81196e+04 5.50667e+05 -2.21641e+04 -4.05276e+06 Coul. recip. Potential Kinetic En. Total Energy Conserved En. 1.15729e+04 -3.38310e+06 5.96289e+05 -2.78681e+06 -2.78768e+06