dptech-corp
CUDA Runtime Error when running Uni-Dock
Hello,
I'm encountering a CUDA Runtime Error when trying to run Uni-Dock on my system. The error message is as follows:
Performing docking (random seed: 1521015319) ... CUDA error at /root/code/Uni-Dock-private/src/cuda/monte_carlo.cu:1719 code=2(cudaErrorMemoryAllocation) "cudaMallocHost(&p_data, sizeof(p_m_data_cuda_t) * MAX_P_DATA_M_DATA_SIZE)"
This error occurs when I run the following command:
./unidock --receptor 7s68_fixed_pH_7.pdbqt --gpu_batch ./ex.pdbqt --search_mode balance --scoring vina --center_x 2.0 --center_y 30.0 --center_z 64.0 --size_x 18.0 --size_y 16.0 --size_z 16.0 --num_modes 20 --dir ./docked
My system is running Windows 10 with WSL2 of Ubuntu 20.04. I have installed the CUDA toolkit (WSL version). When running Uni-Dock, it seems that the GPU is not being utilized, but the CPU is working at 100% load.
Here is the output of nvidia-smi on my system:
Sun Jul 16 20:40:02 2023
+---------------------------------------------------------------------------------------+
| NVIDIA-SMI 535.54.04 Driver Version: 536.23 CUDA Version: 12.2 |
|-----------------------------------------+----------------------+----------------------+
| GPU Name Persistence-M | Bus-Id Disp.A | Volatile Uncorr. ECC |
| Fan Temp Perf Pwr:Usage/Cap | Memory-Usage | GPU-Util Compute M. |
| | | MIG M. |
|=========================================+======================+======================|
| 0 NVIDIA GeForce RTX 4090 On | 00000000:01:00.0 On | Off |
| 0% 41C P8 30W / 450W | 2769MiB / 24564MiB | 11% Default |
| | | N/A |
+-----------------------------------------+----------------------+----------------------+
+---------------------------------------------------------------------------------------+
| Processes: |
| GPU GI CI PID Type Process name GPU Memory |
| ID ID Usage |
|=======================================================================================|
| No running processes found |
+---------------------------------------------------------------------------------------+
Any help in resolving this issue would be greatly appreciated. Thank you!
Add： root@DESKTOP-1IV2ND2:~/screening# nvcc --version nvcc: NVIDIA (R) Cuda compiler driver Copyright (c) 2005-2023 NVIDIA Corporation Built on Fri_Jan__6_16:45:21_PST_2023 Cuda compilation tools, release 12.0, V12.0.140 Build cuda_12.0.r12.0/compiler.32267302_0
Hi @hotwa , Looks like you and @XingyuLiao in #12 are facing the same problem. We are trying to build a WSL environment and further investigate this issue.
execute in docker images： I use this images: nvidia/cuda:latest. root@fa5ddca634eb:/# ./unidock --receptor ./input_file_example/2bm2_protein.pdbqt --gpu_batch ./input_file_example/2bm2_ ligand.pdbqt --search_mode balance --scoring vina --center_x 40.415 --center_y 110.986 --center_z 82.673 --size_x 25 --size_y 25 --size_z 25 --num_modes 20 --dir ./input_file_example/ Uni-Dock v0.1.0
If you used Uni-Dock in your work, please cite:
Yu, Y., Cai, C., Wang, J., Bo, Z., Zhu, Z., & Zheng, H. (2023). Uni-Dock: GPU-Accelerated Docking Enables Ultralarge Virtual Screening. Journal of Chemical Theory and Computation. https://doi.org/10.1021/acs.jctc.2c01145
Tang, S., Chen, R., Lin, M., Lin, Q., Zhu, Y., Ding, J., ... & Wu, J. (2022). Accelerating autodock vina with gpus. Molecules, 27(9), 3041. DOI 10.3390/molecules27093041
J. Eberhardt, D. Santos-Martins, A. F. Tillack, and S. Forli AutoDock Vina 1.2.0: New Docking Methods, Expanded Force Field, and Python Bindings, J. Chem. Inf. Model. (2021) DOI 10.1021/acs.jcim.1c00203
O. Trott, A. J. Olson, AutoDock Vina: improving the speed and accuracy of docking with a new scoring function, efficient optimization and multithreading, J. Comp. Chem. (2010) DOI 10.1002/jcc.21334
Please refer to https://github.com/dptech-corp/Uni-Dock/ for bug reporting, license agreements, and more information.
Scoring function : vina Rigid receptor: ./input_file_example/2bm2_protein.pdbqt Grid center: X 40.415 Y 110.986 Z 82.673 Grid size : X 25 Y 25 Z 25 Grid space : 0.375 Exhaustiveness: 384 CPU: 0 Verbosity: 1
Computing Vina grid ... entering done done. exiting done Total ligands: 1 Avaliable Memory = 23006MiB Total Memory = 24563MiB Batch 1 size: 1 Performing docking (random seed: -1212963753) ... CUDA error at /root/code/Uni-Dock-private/src/cuda/monte_carlo.cu:1719 code=2(cudaErrorMemoryAllocation) "cudaMallocHost(&p_data, sizeof(p_m_data_cuda_t) * MAX_P_DATA_M_DATA_SIZE)"
An error occurred: CUDA Runtime Error.
Please report bugs through the Issue Tracker on GitHub (https://github.com/dptech-corp/Uni-Dock/issues)., so that this problem can be resolved. The reproducibility of the error may be vital, so please remember to include the following in your problem report:
- the EXACT error message,
- your version of the program,
- the type of computer system you are running it on,
- all command line options,
- configuration file (if used),
- ligand file as PDBQT,
- receptor file as PDBQT,
- flexible side chains file as PDBQT (if used),
- output file as PDBQT (if any),
- input (if possible),
- random seed the program used (this is printed when the program starts).
Thank you!
I encountered an issue while using Uni-Dock for molecular docking. When I ran the command ./unidock --receptor ./input_file_example/2bm2_protein.pdbqt --gpu_batch ./input_file_example/2bm2_ligand.pdbqt --search_mode balance --scoring vina --center_x 40.415 --center_y 110.986 --center_z 82.673 --size_x 25 --size_y 25 --size_z 25 --num_modes 20 --dir ./input_file_example, the output showed "Total ligands: 1," but then I received an error message saying "ERROR: Empty ligand list."
Here is my environment information:
Operating System: Ubuntu 20.04 GPU: NVIDIA Titan V Driver Version: 530.30.02 CUDA Version: 12.1 I have confirmed that the ligand file exists and the correct path is specified. I have tried other command and parameter combinations in the same environment, but I still encounter the same issue.
I would like to understand why this error occurs and find a solution. Any help or insights would be greatly appreciated. Thank you!
./unidock --receptor ./input_file_example/2bm2_protein.pdbqt --gpu_batch ./input_file_example/2bm2_ligand.pdbqt --search_mode balance --scoring vina --center_x 40.415 --center_y 110.986 --center_z 82.673 --size_x 25 --size_y 25 --size_z 25 --num_modes 20 --dir ./input_file_example
Uni-Dock v0.1.0
If you used Uni-Dock in your work, please cite:
Yu, Y., Cai, C., Wang, J., Bo, Z., Zhu, Z., & Zheng, H. (2023).
Uni-Dock: GPU-Accelerated Docking Enables Ultralarge Virtual Screening.
Journal of Chemical Theory and Computation.
https://doi.org/10.1021/acs.jctc.2c01145
Tang, S., Chen, R., Lin, M., Lin, Q., Zhu, Y., Ding, J., ... & Wu, J. (2022).
Accelerating autodock vina with gpus. Molecules, 27(9), 3041.
DOI 10.3390/molecules27093041
J. Eberhardt, D. Santos-Martins, A. F. Tillack, and S. Forli
AutoDock Vina 1.2.0: New Docking Methods, Expanded Force
Field, and Python Bindings, J. Chem. Inf. Model. (2021)
DOI 10.1021/acs.jcim.1c00203
O. Trott, A. J. Olson,
AutoDock Vina: improving the speed and accuracy of docking
with a new scoring function, efficient optimization and
multithreading, J. Comp. Chem. (2010)
DOI 10.1002/jcc.21334
Please refer to https://github.com/dptech-corp/Uni-Dock/ for
bug reporting, license agreements, and more information.
Scoring function : vina
Rigid receptor: ./input_file_example/2bm2_protein.pdbqt
Grid center: X 40.415 Y 110.986 Z 82.673
Grid size : X 25 Y 25 Z 25
Grid space : 0.375
Exhaustiveness: 384
CPU: 0
Verbosity: 1
Computing Vina grid ... entering done
done.
exiting done
Total ligands: 1
Avaliable Memory = 4450MiB Total Memory = 12054MiB
Batch 1 size: 0
ERROR: Empty ligand list.
Hi @hotwa , Sorry for the late reply. It seems that your GPU memory size is insufficient for Uni-Dock to launch. We are working on the compatibility on multiple GPUs.
I need at least 24GB of GPU memory for Uni-Dock to launch？
@hotwa 12GB is recommended. Please see https://github.com/dptech-corp/Uni-Dock/issues/9#issuecomment-1616488315
Hi @hotwa ,
I compiled Uni-Dock from source code and successfully finished the task under examples/screening_test. I use ubuntu 22.04 with WSL2, on my 3070 Ti 8GB GPU. It seems that Uni-Dock works smoothly on WSL2 systems.
You could compiling from source code and try again.
When compiled it, I have some trouble. if unidock runing in anaconda, will be great.
When compiled it, I have some trouble. if unidock runing in anaconda, will be great.
Please try:
conda install -c https://conda.mlops.dp.tech/caic -c conda-forge unidock