"Resource Exhausted" error while running lmp

[jaejae9804]

Hi everyone,
greatly appreciate in advance for everyone. What I was doing is running LAMMPS with a Cu dp model developed earlier and encoutered an error metioning what seems to be related to the memory usage. Ive attached the error log, with the input scripts Ive used.
Im using Geforce rtx 4090 as my local computer, and the environment Im using to run deepmd-kit was created with nvidia-docker.
To be specific, the environment Im using to run deepmd-kit is what follows:

Ubuntu-22.04.3 LTS
Geforce rtx 4090
nvidia-docker
docker image: deepmodeling/deepmd-kit:2.2.7_cuda11.6_gpu
(docker container run -it -v /home/jaejae:/root --name dp_2.7 --gpus all deepmodeling/deepmd-kit:2.2.7_cuda11.6_gpu)

If checking out the structural file attached below, you can see that Ive used 100,0188 Cu atoms for LAMMPS testing. When running the same simulation with the identical input scripts with 4000atoms, everything seems to be fine, meaning that none of the following errors happened. It would be of a great help if anyone can give an advice on how to solve this problem. Thanks.
error log.txt
in.lammps.txt
xo.lmp.txt

[njzjz]

You can use the MPI to run on multiple GPU cards if one card doesn't have enough memory.

[jaejae9804]

Thank you for the response. Unfortunately, I do have only one gpu card in my local computer...If in my case, does mpirun have benefit in fixing this problem?

[jaejae9804]

Oh, just in case Ive been less specific, the memory of the gpu card is 32GB, and the additional infos about the cpus are as follows:

jaejae@DESKTOP-ORMR4QC:~$ free
total used free shared buff/cache available
Mem: 32707244 2529324 28396780 347484 1781140 29428268
Swap: 8388608 0 8388608

jaejae@DESKTOP-ORMR4QC:~/Sn/dpgen/data/b763195f5224117522baa37fbf29673885f91154/003$ lscpu
Architecture: x86_64
CPU op-mode(s): 32-bit, 64-bit
Address sizes: 46 bits physical, 48 bits virtual
Byte Order: Little Endian
CPU(s): 28
On-line CPU(s) list: 0-27
Vendor ID: GenuineIntel
Model name: Intel(R) Core(TM) i7-14700K
CPU family: 6
Model: 183
Thread(s) per core: 2
Core(s) per socket: 14
Socket(s): 1
Stepping: 1

[njzjz]

To run 1 million atoms with a compressed model, about 60GB of GPU memory is required. Please see Table 5 at https://doi.org/10.1063/5.0155600.