simtk.openmm.OpenMMException: Error loading CUDA module: CUDA_ERROR_INVALID_PTX (218) about seekr2 HOT 14 CLOSED

Dear Vaibhav,
I think it would be useful to know how you have installed seekr2? Especially, if you used conda or compiled from source.

I believe, the seekr2-openmm-plugin which is accessible by conda-forge comes precompiled against cuda10.2. Perhaps this is already the issue but Lane is the expert here.

Best regards
Jens Wilbert

from seekr2.

Dear Jens,
OK, I think I'm able to understand the incompatibility that you've described.
I created a SEEKR environment in conda, then installed python3.8, then used conda to install the seekr2-openmm-plugin and seekr2. (FYI - Lane is already aware of my installation failure/success attempts, thus I'll also wait for him to suggest further things to try).
thanks and best regards.
Vaibhav

from seekr2.

from seekr2.

Hi Anand,
thanks for your response.
I have earlier interacted with Lane, who recommended creating a SEEKR environment in python3.8 and then calling openmm modules. Below is the usage that is currently showing that it is installed as per suggestions.
Please let me know additional info about the current installation is required.
thanks
(base) [exx@c107739 ~]$ conda activate SEEKR
(SEEKR) [exx@c107739 ~]$ python3.8
Python 3.8.10 | packaged by conda-forge | (default, May 11 2021, 07:01:05)
[GCC 9.3.0] on linux
Type "help", "copyright", "credits" or "license" for more information.

from simtk import openmm

from seekr2.

from seekr2.

Hi Anand,
Your suspicion was correct, I'm getting the following which shows that cuda gives error.
Please do let me know what you think would be the best options to try now.
thanks.

(base) [exx@c107739 ~]$ conda activate SEEKR
(SEEKR) [exx@c107739 ~]$ python -m simtk.testInstallation

OpenMM Version: 7.5.1
Git Revision: a9cfd7fb9343e21c3dbb76e377c721328830a3ee

There are 4 Platforms available:

1 Reference - Successfully computed forces
2 CPU - Successfully computed forces
3 CUDA - Error computing forces with CUDA platform
4 OpenCL - Successfully computed forces

CUDA platform error: Error loading CUDA module: CUDA_ERROR_INVALID_PTX (218)

Median difference in forces between platforms:

Reference vs. CPU: 6.29101e-06
Reference vs. OpenCL: 6.75475e-06
CPU vs. OpenCL: 8.07327e-07

All differences are within tolerance.
(SEEKR) [exx@c107739 ~]$

from seekr2.

Hello,

Thanks @anandojha and @bensebaer for your comments and assistance.

Yes, as the linked OpenMM issue suggested, this error is caused by a driver incompatibility or also an incompatible GPU. Do you know what type of GPU you are using and whether it is compatible with CUDA 10.2? (The command "nvidia-smi" might work, and it will tell you your driver versions, CUDA version, and what GPUs you have).

I noticed from your first post that the nvcc version you are using is for CUDA 10.1, which is not compatible with the latest OpenMM and, by extension, SEEKR2. How did you install that version? Was that CUDA build from your previous attempts at installation? One solution would be to uninstall your CUDA 10.1, then retry the SEEKR2 installation, and Conda should at least pull the correct version of CudaTools. Or you could try to install CUDA 10.2 manually. The best choice depends on your situation.

from seekr2.

Dear Lane,
I have RTX2080 GPU hardware as shown below.
Is it possible to install cuda 10.2 in the seekr environment without affecting the older version on which my current Amber18 installation depends?
If this is possible then I will do that and try to reinstall seekr2.
Please let me know.
thanks

(base) [user@localhost ~]$ nvidia-smi
Wed Sep 1 20:58:36 2021
+-----------------------------------------------------------------------------+
| NVIDIA-SMI 440.33.01 Driver Version: 440.33.01 CUDA Version: 10.2 |
|-------------------------------+----------------------+----------------------+
| GPU Name Persistence-M| Bus-Id Disp.A | Volatile Uncorr. ECC |
| Fan Temp Perf Pwr:Usage/Cap| Memory-Usage | GPU-Util Compute M. |
|===============================+======================+======================|
| 0 GeForce RTX 2080 Off | 00000000:65:00.0 Off | N/A |
| 0% 40C P8 5W / 225W | 1MiB / 7982MiB | 0% Default |
+-------------------------------+----------------------+----------------------+
| 1 GeForce RTX 2080 Off | 00000000:B3:00.0 Off | N/A |
| 0% 41C P8 1W / 225W | 358MiB / 7981MiB | 0% Default |
+-------------------------------+----------------------+----------------------+

+-----------------------------------------------------------------------------+
| Processes: GPU Memory |
| GPU PID Type Process name Usage |
|=============================================================================|
| 1 2363 G /usr/bin/X 148MiB |
| 1 3233 G /usr/bin/gnome-shell 124MiB |
| 1 56760 G ...AAAAAAAAAAAACAAAAAAAAAA= --shared-files 83MiB |
+-----------------------------------------------------------------------------+
(base) [user@localhost ~]$

from seekr2.

OK thanks for that information. Yes, your RTX2080s should work just fine, and the driver version looks recent enough. Your Nvidia-smi command shows that it thinks you have Cuda10.2. Do you have multiple Cuda installations? You should look in /usr/local/ to see.

There may be some possible options for you:

1. You could try to install (or find, if it's already installed) CUDA 10.2 (or later) in addition to CUDA10.1, and then set the OPENMM_CUDA_COMPILER environmental variable to point to the CUDA 10.2 nvcc path, but somehow point Amber18 to the 10.1 version (probably another environmental variable).
2. You might be able to install an older version of OpenMM that was compatible with Cuda10.1 and then install SEEKR2 on top of that (probably not a good idea, and you'd probably need to install the seekr2 openmm plugin from source on top of it).
3. You might be able to make the SEEKR conda environment ignore the installed CUDA by setting the CUDA_HOME and CUDA_ROOT environmental variables to empty for the environment (other variables might need to be set to empty as well), then when you install the seekr2 openmm plugin, it should install the proper version of CudaToolkit.

from seekr2.

Hi,
I tried the first option suggested by Lane and installed cuda 10.2 as suggested here and nvcc --version returns cuda 10.2.
But I'm still getting driver errors. Can you please check and suggest to me what to try next?

https://www.server-world.info/en/note?os=CentOS_7&p=cuda&f=4

(SEEKR) [exx@c107739 seekr2]$ nvcc --version
nvcc: NVIDIA (R) Cuda compiler driver
Copyright (c) 2005-2019 NVIDIA Corporation
Built on Wed_Oct_23_19:24:38_PDT_2019
Cuda compilation tools, release 10.2, V10.2.89
(SEEKR) [exx@c107739 seekr2]$

(SEEKR) [exx@c107739 seekr2]$ python3.8 run.py any ~/test_mmvt_openmm/model.xml
anchor 0 has not run the minimum number of steps 0 of 100000
anchor 1 has not run the minimum number of steps 0 of 100000
anchor 2 has not run the minimum number of steps 0 of 100000
anchor 3 has not run the minimum number of steps 0 of 100000
anchor 4 has not run the minimum number of steps 0 of 100000
anchor 5 has not run the minimum number of steps 0 of 100000
anchor 6 has not run the minimum number of steps 0 of 100000
anchor 7 has not run the minimum number of steps 0 of 100000
anchor 8 has not run the minimum number of steps 0 of 100000
anchor 9 has not run the minimum number of steps 0 of 100000
anchor 10 has not run the minimum number of steps 0 of 100000
anchor 11 has not run the minimum number of steps 0 of 100000
anchor 12 has not run the minimum number of steps 0 of 100000
running anchor_index: 0 restart: False total_simulation_length: 100000 num_transitions: 0
Traceback (most recent call last):
File "run.py", line 742, in
run(model, instruction, min_total_simulation_length,
File "run.py", line 515, in run
run_openmm(model, anchor_index, restart,
File "run.py", line 404, in run_openmm
sim_openmm_obj = mmvt_sim_openmm.create_sim_openmm(
File "/home/exx/anaconda3/envs/SEEKR/lib/python3.8/site-packages/seekr2-0+untagged.144.ge74fc0f-py3.8.egg/seekr2/modules/mmvt_sim_openmm.py", line 168, in create_sim_openmm
add_simulation(sim_openmm, model, topology, positions, box_vectors)
File "/home/exx/anaconda3/envs/SEEKR/lib/python3.8/site-packages/seekr2-0+untagged.144.ge74fc0f-py3.8.egg/seekr2/modules/mmvt_sim_openmm.py", line 101, in add_simulation
sim_openmm.simulation = openmm_app.Simulation(
File "/home/exx/anaconda3/envs/SEEKR/lib/python3.8/site-packages/simtk/openmm/app/simulation.py", line 105, in init
self.context = mm.Context(self.system, self.integrator, platform, platformProperties)
File "/home/exx/anaconda3/envs/SEEKR/lib/python3.8/site-packages/simtk/openmm/openmm.py", line 13232, in init
_openmm.Context_swiginit(self, _openmm.new_Context(*args))
simtk.openmm.OpenMMException: Error initializing CUDA: CUDA_ERROR_SYSTEM_DRIVER_MISMATCH (803) at /home/conda/feedstock_root/build_artifacts/openmm_1622798691062/work/platforms/cuda/src/CudaContext.cpp:138
(SEEKR) [exx@c107739 seekr2]$

from seekr2.

Hi SEEKR team,
To my surprise, the calculations seem to be running just fine (after there was a reboot due to power failure) and I accidentally decided to run the tests/tutorial case anyway.
Can you please have a look and let me know if it is doing things as expected?
Can you please suggest a set of commands that can print the full setup on which SEEKR2 is depending here?
This will help me try to reproduce the installation on another GPU similar system that I have in the lab.
I wasn't expecting a reboot to fix something for me here (which can often happen in windows OS).
Did I just got lucky here due to some OS update in the background or I did something to fix it without realizing it myself?
Looking forward to your confirmation and suggestions.
thank you and best regards.

(SEEKR) [exx@c107739 seekr2]$ python3.8 run.py any ~/test_mmvt_openmm/model.xml
anchor 0 has not run the minimum number of steps 0 of 100000
anchor 1 has not run the minimum number of steps 0 of 100000
anchor 2 has not run the minimum number of steps 0 of 100000
anchor 3 has not run the minimum number of steps 0 of 100000
anchor 4 has not run the minimum number of steps 0 of 100000
anchor 5 has not run the minimum number of steps 0 of 100000
anchor 6 has not run the minimum number of steps 0 of 100000
anchor 7 has not run the minimum number of steps 0 of 100000
anchor 8 has not run the minimum number of steps 0 of 100000
anchor 9 has not run the minimum number of steps 0 of 100000
anchor 10 has not run the minimum number of steps 0 of 100000
anchor 11 has not run the minimum number of steps 0 of 100000
anchor 12 has not run the minimum number of steps 0 of 100000
running anchor_index: 0 restart: False total_simulation_length: 100000 num_transitions: 0
#"Step","Potential Energy (kJ/mole)","Temperature (K)","Box Volume (nm^3)"
10000,-69144.49079205882,297.4233782909998,52.56848950689602
20000,-69869.93366379052,301.560597952733,52.56848950689602
30000,-69864.64904012525,296.51825589367263,52.56848950689602
40000,-69377.52449393808,302.80793589861304,52.56848950689602
50000,-69721.83292784536,292.03861392671246,52.56848950689602
60000,-70285.15553236526,298.7569602733776,52.56848950689602
70000,-69477.06213368417,302.12781557852315,52.56848950689602
80000,-69412.19511798577,298.04278694035133,52.56848950689602
90000,-69802.81410196872,302.73667807143323,52.56848950689602
100000,-69872.02696706454,300.25169928173386,52.56848950689602
Benchmark (ns/day): 676.0865219728388
running anchor_index: 1 restart: False total_simulation_length: 100000 num_transitions: 0
#"Step","Potential Energy (kJ/mole)","Temperature (K)","Box Volume (nm^3)"
10000,-69701.0679280666,299.9465145748033,53.001182076839996
20000,-68875.80444348347,291.08401098832985,53.001182076839996
30000,-69407.27905209706,295.04682130819543,53.001182076839996
40000,-69374.98674597323,296.5943957999594,53.001182076839996
50000,-69510.50046643266,297.4515442928506,53.001182076839996
60000,-69487.54161233205,300.2598541155219,53.001182076839996
70000,-69621.41305995733,300.3923321056398,53.001182076839996
80000,-69524.28550882259,303.74551240996215,53.001182076839996
90000,-70045.26211276255,311.0443026811395,53.001182076839996
100000,-69362.11746455648,296.17904476577297,53.001182076839996
Benchmark (ns/day): 661.0541200182117
running anchor_index: 2 restart: False total_simulation_length: 100000 num_transitions: 0
#"Step","Potential Energy (kJ/mole)","Temperature (K)","Box Volume (nm^3)"
10000,-69785.76114975323,296.46893362002567,53.001182076839996
20000,-69200.05805598001,296.41330462774545,53.001182076839996
30000,-69295.07416729943,302.4301459305852,53.001182076839996
40000,-69281.83241042274,301.9972234557154,53.001182076839996
50000,-69091.48401152418,295.9802550223431,53.001182076839996
60000,-69381.06063582585,294.0204234760598,53.001182076839996
70000,-69486.26055974478,298.1119611029692,53.001182076839996
80000,-69115.26048954594,299.44512274838866,53.001182076839996
90000,-69300.33501393104,296.359589787676,53.001182076839996
100000,-69486.955681903,307.9725501983454,53.001182076839996
Benchmark (ns/day): 656.2592886141096
running anchor_index: 3 restart: False total_simulation_length: 100000 num_transitions: 0
#"Step","Potential Energy (kJ/mole)","Temperature (K)","Box Volume (nm^3)"
10000,-69737.18051548884,294.5279735085756,53.001182076839996
20000,-69392.46493441012,293.44891176706403,53.001182076839996
30000,-69498.79043459706,303.006132158916,53.001182076839996
40000,-69175.29168446176,294.6025852434763,53.001182076839996
50000,-69383.68013051013,306.0325861102716,53.001182076839996
60000,-69961.78327824437,294.75652817405745,53.001182076839996
70000,-69871.04464263469,299.42534139155663,53.001182076839996
80000,-69456.92276385019,300.3888699871004,53.001182076839996
90000,-69786.59675886354,298.23651743435175,53.001182076839996
100000,-69643.17584280891,298.11451407893713,53.001182076839996
Benchmark (ns/day): 614.2356845813468
running anchor_index: 4 restart: False total_simulation_length: 100000 num_transitions: 0
#"Step","Potential Energy (kJ/mole)","Temperature (K)","Box Volume (nm^3)"
10000,-69386.6953338632,303.95054076402243,53.001182076839996
20000,-69253.00790108804,300.47427564619716,53.001182076839996
30000,-69465.83993170381,292.36135086040895,53.001182076839996
40000,-69448.45238402707,295.603624863649,53.001182076839996
50000,-69452.6658732178,305.1601520774961,53.001182076839996
60000,-69162.39610835159,300.0348418201349,53.001182076839996
70000,-69205.13485659898,291.70067787432737,53.001182076839996
80000,-69333.56733806909,293.75808921470457,53.001182076839996
90000,-69219.3913178423,298.9393149480047,53.001182076839996
100000,-69191.19156954251,295.99066890970664,53.001182076839996
Benchmark (ns/day): 617.8916371039779
running anchor_index: 5 restart: False total_simulation_length: 100000 num_transitions: 0
#"Step","Potential Energy (kJ/mole)","Temperature (K)","Box Volume (nm^3)"
10000,-69457.26372874231,298.5604264331683,53.001182076839996
20000,-69605.07306833827,304.36473799555705,53.001182076839996
30000,-69276.33479454223,299.89061760295914,53.001182076839996
40000,-69371.19127700396,298.2033213307406,53.001182076839996
50000,-69446.38602255401,299.05456112032755,53.001182076839996
60000,-68944.29696569016,291.75485399398684,53.001182076839996
70000,-69678.53721238917,299.12353092231524,53.001182076839996
80000,-69207.1422912617,293.6369841531982,53.001182076839996
90000,-69158.74460100941,294.2481936769401,53.001182076839996
100000,-69736.93902522675,300.350724483127,53.001182076839996
Benchmark (ns/day): 635.8922623463477
running anchor_index: 6 restart: False total_simulation_length: 100000 num_transitions: 0
#"Step","Potential Energy (kJ/mole)","Temperature (K)","Box Volume (nm^3)"
10000,-69251.31409991405,300.405627763369,53.001182076839996
20000,-69504.57336028386,303.30762338068723,53.001182076839996
30000,-69881.59594982542,297.0480543670246,53.001182076839996
40000,-69196.91975443647,296.54985897734304,53.001182076839996
50000,-69337.93296267267,302.9833374553011,53.001182076839996
60000,-69786.15843727044,298.66235433774364,53.001182076839996
70000,-69565.88825172913,310.51878997243443,53.001182076839996
80000,-69561.05699932424,296.2620756755379,53.001182076839996
90000,-68921.38426727621,293.9531165881675,53.001182076839996
100000,-69359.30978567398,298.37334101205585,53.001182076839996
Benchmark (ns/day): 640.9204718489292
running anchor_index: 7 restart: False total_simulation_length: 100000 num_transitions: 0
#"Step","Potential Energy (kJ/mole)","Temperature (K)","Box Volume (nm^3)"
10000,-69362.65660170786,298.4715200397846,53.001182076839996
20000,-69459.3173377564,294.9194907853032,53.001182076839996
30000,-69137.4052030241,305.3206715784008,53.001182076839996
40000,-69733.47936457687,298.550087659625,53.001182076839996
50000,-69206.60384401074,304.203192377914,53.001182076839996
60000,-69728.44668358349,302.77903815229723,53.001182076839996
70000,-69640.23172272567,296.61623855379094,53.001182076839996
80000,-69180.34710867377,298.3687018439519,53.001182076839996
90000,-69913.56008797663,296.9220362845659,53.001182076839996
100000,-69325.02275857137,290.20203682225156,53.001182076839996
Benchmark (ns/day): 621.8316666762902
running anchor_index: 8 restart: False total_simulation_length: 100000 num_transitions: 0
#"Step","Potential Energy (kJ/mole)","Temperature (K)","Box Volume (nm^3)"
10000,-68739.53935059876,289.42954985552257,53.001182076839996
20000,-69121.00007149944,299.9123223974211,53.001182076839996
30000,-69811.68617732049,296.477888605543,53.001182076839996
40000,-69093.4523969164,295.3691913464193,53.001182076839996
50000,-69373.34382889944,299.4915681131448,53.001182076839996
60000,-69568.40951526398,308.6268661829777,53.001182076839996
70000,-69318.72763755795,295.679862981259,53.001182076839996
80000,-69550.27482223621,294.32184033944105,53.001182076839996
90000,-69380.45023123675,296.3258739871022,53.001182076839996
100000,-69723.430675623,289.9096700042352,53.001182076839996
Benchmark (ns/day): 648.3124859183208
running anchor_index: 9 restart: False total_simulation_length: 100000 num_transitions: 0
#"Step","Potential Energy (kJ/mole)","Temperature (K)","Box Volume (nm^3)"
10000,-69858.92770797084,291.9108298301936,53.001182076839996
20000,-69792.74257540028,296.94819340770266,53.001182076839996
30000,-69147.37162053975,306.5790999340041,53.001182076839996
40000,-69840.14387600496,298.09405670187914,53.001182076839996
50000,-69124.67333274201,290.6085592263482,53.001182076839996
60000,-69837.37839602039,292.0942978063598,53.001182076839996
70000,-69220.87221926934,292.0224410766485,53.001182076839996
80000,-69529.60648475884,303.9038645638152,53.001182076839996
90000,-69695.44364100188,303.63866094972843,53.001182076839996
100000,-69796.55966087629,305.4785144705075,53.001182076839996
Benchmark (ns/day): 618.3341983021858
running anchor_index: 10 restart: False total_simulation_length: 100000 num_transitions: 0
#"Step","Potential Energy (kJ/mole)","Temperature (K)","Box Volume (nm^3)"
10000,-68734.58490729053,290.97723757808507,53.001182076839996
20000,-69255.29286105372,299.26811022468314,53.001182076839996
30000,-69180.11697052879,302.9003978966918,53.001182076839996
40000,-69909.11901671207,306.9346781786293,53.001182076839996
50000,-69607.68029917893,303.12062335102496,53.001182076839996
60000,-69131.29972944304,300.0868214441313,53.001182076839996
70000,-69479.25102394383,296.8862636112948,53.001182076839996
80000,-69484.68777931464,298.91813823296667,53.001182076839996
90000,-69448.93743409216,296.58130712262135,53.001182076839996
100000,-69244.11212478636,299.0142822728575,53.001182076839996
Benchmark (ns/day): 620.5240910663637
running anchor_index: 11 restart: False total_simulation_length: 100000 num_transitions: 0
#"Step","Potential Energy (kJ/mole)","Temperature (K)","Box Volume (nm^3)"
10000,-69604.82230700646,291.95329672997866,53.001182076839996
20000,-69510.67855428415,295.80811020616716,53.001182076839996
30000,-69396.76138013112,299.307601813386,53.001182076839996
40000,-70003.28882299172,303.74270547547115,53.001182076839996
50000,-69500.36680087273,303.5304438083032,53.001182076839996
60000,-69442.89330834162,291.56155136654564,53.001182076839996
70000,-69363.14966853557,301.3103345014555,53.001182076839996
80000,-69963.57409431,298.82918600416957,53.001182076839996
90000,-69208.87962260656,301.6396460181288,53.001182076839996
100000,-69553.3454008653,298.0939068570571,53.001182076839996
Benchmark (ns/day): 621.1309564011887
running anchor_index: 12 restart: False total_simulation_length: 100000 num_transitions: 0
#"Step","Potential Energy (kJ/mole)","Temperature (K)","Box Volume (nm^3)"
10000,-69175.02726903511,296.3507783306151,53.001182076839996
20000,-69305.08690150117,300.6441185317881,53.001182076839996
30000,-69498.77728664369,301.5865100771034,53.001182076839996
40000,-69378.32120181562,296.7177506805017,53.001182076839996
50000,-69597.09609594557,296.0285238252822,53.001182076839996
60000,-69285.1623312519,298.64573238303086,53.001182076839996
70000,-69805.5251628966,294.6759730709682,53.001182076839996
80000,-69298.1716454933,300.49698663034195,53.001182076839996
90000,-69851.00703219848,299.06179501114514,53.001182076839996
100000,-69256.60992451839,298.7163790259338,53.001182076839996
Benchmark (ns/day): 643.8215685422085
running BD: b_surface restart: False trajectories to run: 10000 transitions so far: 0
moving to directory: /home/exx/test_mmvt_openmm/b_surface
running command: bd_top input.xml
command surface_spheres -probe_radius 1.5 < hostguest_receptor.xml > receptor_surface.xml
command inside_points -spheres hostguest_receptor.xml -surface receptor_surface.xml -egrid receptor2.dx > receptor_inside.xml
command hydro_params < receptor_inside.xml > receptor_hydro_params.xml
command test_charges < hostguest_receptor.xml > receptor_charges.xml
command lumped_charges -pts receptor_charges.xml > receptor_cheby.xml
command make_surface_sphere_list -surface receptor_surface.xml -spheres hostguest_receptor.xml -rxn rxns.xml -group receptor -core receptor > receptor_surface_atoms.xml
command mpole_grid_fit -dx receptor0.dx -solvdi 78 -debye 1.79769e+308 > receptor_mpole.xml
command born_integral -in receptor_inside.xml -oeps 78 -debye 1.79769e+308 -ieps 4.000000 > receptor_born.dx
command compute_charges_squared < receptor_charges.xml > receptor_charges_squared.xml
command lumped_charges -pts receptor_charges_squared.xml > receptor_squared_cheby.xml
command surface_spheres -probe_radius 1.5 < hostguest_ligand.xml > ligand_surface.xml
command inside_points -spheres hostguest_ligand.xml -surface ligand_surface.xml -egrid ligand2.dx > ligand_inside.xml
command hydro_params < ligand_inside.xml > ligand_hydro_params.xml
command test_charges < hostguest_ligand.xml > ligand_charges.xml
command lumped_charges -pts ligand_charges.xml > ligand_cheby.xml
command make_surface_sphere_list -surface ligand_surface.xml -spheres hostguest_ligand.xml -rxn rxns.xml -group ligand -core ligand > ligand_surface_atoms.xml
command mpole_grid_fit -dx ligand0.dx -solvdi 78 -debye 1.79769e+308 > ligand_mpole.xml
command born_integral -in ligand_inside.xml -oeps 78 -debye 1.79769e+308 -ieps 4.000000 > ligand_born.dx
command compute_charges_squared < ligand_charges.xml > ligand_charges_squared.xml
command lumped_charges -pts ligand_charges_squared.xml > ligand_squared_cheby.xml
moving to directory: /home/exx/test_mmvt_openmm/b_surface
running command: nam_simulation receptor_ligand_simulation.xml
BrownDye 2.0: Version of 21 Jul 2021

from seekr2.

from seekr2.

Hi All and @anandojha,
This is somewhat disappointing because it means that it would take more than standard effort to install it on a new system each time. Do you have a sort of idea/benchmark on differences in results that might arise due to different installation details?
I reran the simtk.testInstallation option, which suggests that CUDA version is working now.
But which output file can I check to know/test if a calculation used the CUDA-GPU?
The calculation is already over thus can't check it with the nvidia-smi option, right?

Regarding the tutorial/test case:
The run step finished without errors and did not print anything additional beyond what I shared in the last message, but.
The next "converge" and analyze steps ran only when I invoked python3.8.
Can you please check if it converged and was analyzed correctly? It looks like it failed more than a couple of times.
Thus do I need to change the input and start thinking about the science (milestoning theory) based changes in the input.
Please suggest.
Looking forward to your valuable suggestions and comments.
thanks and best regards.

(SEEKR) [exx@c107739 seekr2]$ python3.8 -m simtk.testInstallation

OpenMM Version: 7.5.1
Git Revision: a9cfd7fb9343e21c3dbb76e377c721328830a3ee

There are 4 Platforms available:

1 Reference - Successfully computed forces
2 CPU - Successfully computed forces
3 CUDA - Successfully computed forces
4 OpenCL - Successfully computed forces

Median difference in forces between platforms:

Reference vs. CPU: 6.29272e-06
Reference vs. CUDA: 6.74111e-06
CPU vs. CUDA: 7.36849e-07
Reference vs. OpenCL: 6.75475e-06
CPU vs. OpenCL: 8.03983e-07
CUDA vs. OpenCL: 2.67005e-07

All differences are within tolerance.
(SEEKR) [exx@c107739 seekr2]$

(SEEKR) [exx@c107739 seekr2]$ python converge.py ~/test_mmvt_openmm/model.xml
Traceback (most recent call last):
File "converge.py", line 9, in
import argparse
ImportError: No module named argparse
(SEEKR) [exx@c107739 seekr2]$ python3.8 converge.py ~/test_mmvt_openmm/model.xml
All available options for --k_on_state include: [0]
No BD milestone has been provided for k_on_state. The BD milestone 0 has been chosen by default.
Processing interval 0 of 100
Processing interval 10 of 100
Processing interval 20 of 100
Processing interval 30 of 100
Processing interval 40 of 100
Processing interval 50 of 100
Processing interval 60 of 100
Processing interval 70 of 100
Processing interval 80 of 100
Processing interval 90 of 100
All plots have been saved to: /home/exx/test_mmvt_openmm/images_and_plots/
Molecular dynamics results:

• Anchor 0:
  Milestone transitions: 0->0 : 165,
  Milestone avg. transition time (ps): (0, 0) : 1.193,
  Convergence value: 6.0899e-03.
  Converged? True
• Anchor 1:
  Milestone transitions: 0->1 : 16, 1->0 : 16,
  Milestone avg. transition time (ps): 0 : 8.382, 1 : 3.977,
  Convergence value: 1.4208e-02.
  Converged? False
• Anchor 2:
  Milestone transitions: 2->1 : 12, 1->2 : 11,
  Milestone avg. transition time (ps): 2 : 1.119, 1 : 15.319,
  Convergence value: 4.4027e-02.
  Converged? False
• Anchor 3:
  Milestone transitions: 2->3 : 25, 3->2 : 24,
  Milestone avg. transition time (ps): 2 : 3.560, 3 : 4.354,
  Convergence value: 1.5193e-02.
  Converged? False
• Anchor 4:
  Milestone transitions: 3->4 : 30, 4->3 : 30,
  Milestone avg. transition time (ps): 3 : 4.569, 4 : 1.976,
  Convergence value: 1.4800e-02.
  Converged? False
• Anchor 5:
  Milestone transitions: 5->4 : 39, 4->5 : 39,
  Milestone avg. transition time (ps): 5 : 1.756, 4 : 3.271,
  Convergence value: 2.1740e-02.
  Converged? False
• Anchor 6:
  Milestone transitions: 6->5 : 24, 5->6 : 23,
  Milestone avg. transition time (ps): 6 : 4.478, 5 : 3.476,
  Convergence value: 9.6712e-03.
  Converged? False
• Anchor 7:
  Milestone transitions: 7->6 : 23, 6->7 : 22,
  Milestone avg. transition time (ps): 7 : 4.550, 6 : 4.079,
  Convergence value: 6.3605e-03.
  Converged? False
• Anchor 8:
  Milestone transitions: 8->7 : 29, 7->8 : 28,
  Milestone avg. transition time (ps): 8 : 3.782, 7 : 2.884,
  Convergence value: 8.4210e-03.
  Converged? False
• Anchor 9:
  Milestone transitions: 9->8 : 21, 8->9 : 21,
  Milestone avg. transition time (ps): 9 : 5.848, 8 : 3.112,
  Convergence value: 1.1400e-02.
  Converged? False
• Anchor 10:
  Milestone transitions: 9->10 : 19, 10->9 : 18,
  Milestone avg. transition time (ps): 9 : 6.821, 10 : 3.433,
  Convergence value: 9.9480e-03.
  Converged? False
• Anchor 11:
  Milestone transitions: 10->11 : 30, 11->10 : 30,
  Milestone avg. transition time (ps): 10 : 3.122, 11 : 3.429,
  Convergence value: 5.6892e-03.
  Converged? False
• Anchor 12:
  Milestone transitions: 12->11 : 32, 11->12 : 31,
  Milestone avg. transition time (ps): 12 : 4.020, 11 : 2.067,
  Convergence value: 7.0283e-03.
  Converged? False
  Brownian dynamics results:
• b-surface reaction counts:
  to milestone 12: 4817
  to milestone 11: 2404
  to escaped: 7596
  to stuck: 0
• BD milestone 0 reaction counts:
  to milestone 11: 2404
  to escaped: 2413
  (SEEKR) [exx@c107739 seekr2]$ python3.8 analyze.py ~/test_mmvt_openmm/model.xml
  All post-simulation checks passed.
  Printing results from MMVT SEEKR calculation
  k_off (1/s): (1.27 ± 0.18) * 10^+07
  k_ons :
  k_on (1/s * 1/M) to state 0 : (3.570 ± 0.052) * 10^+08
  Dissociation constant (M) to state 0 : (3.56 ± 0.50) * 10^-02
  ΔG (kcal/mol) to state 0 : (-1.976 ± 0.083) * 10^+00
  Mean first passage times (s):
  MFPT from state 0 to state bulk : (7.9 ± 1.0) * 10^-08
  All plots being saved to: /home/exx/test_mmvt_openmm/images_and_plots/
  (SEEKR) [exx@c107739 seekr2]$

from seekr2.

This issue was resolved by restarting the computer.

from seekr2.