Compatibility Issue on `midway3-0372`
yihengwuKP opened this issue · comments
Summary
Running openmm
ABF example script alanine-dipeptide.py
in midway3-0372
would lead to errors, whereas for other nodes under depablo-gpu
, e.g. midway3-0301, midway3-0303, midway3-0299, it ran successfully. Perhaps someone with admin power @ndtrung81 can do more debugging.
Procedure to reproduce
ssh midway3-0372 # I have a job running on it. haven't tested on interactive mode due to node availability
module unload pysages
module unload python
module unload gcc
module load pysages
cd ~/source_code/myPysages/examples/openmm/abf/
python alanine-dipeptide_openmm.py
and it throws:
2024-07-03 17:26:08.092095: W external/xla/xla/stream_executor/gpu/asm_compiler.cc:231] Falling back to the CUDA driver for PTX compilation; ptxas does not support CC 9.0
2024-07-03 17:26:08.092155: W external/xla/xla/stream_executor/gpu/asm_compiler.cc:234] Used ptxas at ptxas
2024-07-03 17:26:08.121780: F external/xla/xla/service/gpu/gpu_executable.cc:322] Check failed: !info.content.empty()
[1] 1637925 abort (core dumped) python alanine-dipeptide_openmm.py
For debugging purpose, here is the output of python -c "import jax; jax.print_environment_info()"
jax: 0.4.13
jaxlib: 0.4.13
numpy: 1.24.4
python: 3.8.17 | packaged by conda-forge | (default, Jun 16 2023, 07:06:00) [GCC 11.4.0]
jax.devices (4 total, 4 local): [gpu(id=0) gpu(id=1) gpu(id=2) gpu(id=3)]
process_count: 1
$ nvidia-smi
Wed Jul 3 17:27:39 2024
+---------------------------------------------------------------------------------------+
| NVIDIA-SMI 535.161.08 Driver Version: 535.161.08 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 H100 PCIe On | 00000000:1A:00.0 Off | 0 |
| N/A 44C P0 79W / 350W | 922MiB / 81559MiB | 12% Default |
| | | Disabled |
+-----------------------------------------+----------------------+----------------------+
| 1 NVIDIA H100 PCIe On | 00000000:1C:00.0 Off | 0 |
| N/A 48C P0 84W / 350W | 922MiB / 81559MiB | 17% Default |
| | | Disabled |
+-----------------------------------------+----------------------+----------------------+
| 2 NVIDIA H100 PCIe On | 00000000:B4:00.0 Off | 0 |
| N/A 79C P0 346W / 350W | 6874MiB / 81559MiB | 99% Default |
| | | Disabled |
+-----------------------------------------+----------------------+----------------------+
| 3 NVIDIA H100 PCIe On | 00000000:B6:00.0 Off | 0 |
| N/A 41C P0 77W / 350W | 922MiB / 81559MiB | 14% Default |
| | | Disabled |
+-----------------------------------------+----------------------+----------------------+
+---------------------------------------------------------------------------------------+
| Processes: |
| GPU GI CI PID Type Process name GPU Memory |
| ID ID Usage |
|=======================================================================================|
| 0 N/A N/A 1608287 C python 454MiB |
| 0 N/A N/A 1638225 C python 454MiB |
| 1 N/A N/A 1608286 C python 454MiB |
| 1 N/A N/A 1638225 C python 454MiB |
| 2 N/A N/A 928404 C python3 5628MiB |
| 2 N/A N/A 1638225 C python 454MiB |
| 3 N/A N/A 1608291 C python 454MiB |
| 3 N/A N/A 1638225 C python 454MiB |
+---------------------------------------------------------------------------------------+