Dear CP2K community,
I am interested to run CP2K using GPU support on LUMI-G HPC.
I encounter a problem while running DFTMD, where the program terminates after approximately 1000 MD steps. Even if I restart the simulation again, ignoring this message, it crashes at another 1000 MD steps, indicating there is some issue with memory processing.
The std error file complains as follows:
ERROR: "hipErrorOutOfMemory" at /pfs/lustrep4/projappl/project_462000007/rkronber/cp2k-2025.1/src/grid/hip/../../offload/offload_runtime.h:379
Program received signal SIGABRT: Process abort signal.
Backtrace for this error:
ERROR: "hipErrorOutOfMemory" at /pfs/lustrep4/projappl/project_462000007/rkronber/cp2k-2025.1/src/grid/hip/../../offload/offload_runtime.h:379
Program received signal SIGABRT: Process abort signal.
Backtrace for this error:
ERROR: "hipErrorOutOfMemory" at /pfs/lustrep4/projappl/project_462000007/rkronber/cp2k-2025.1/src/grid/hip/../../offload/offload_runtime.h:379
Program received signal SIGABRT: Process abort signal.
.
.
.
.
continues with more error statements. I have attached the std error and other files for reference.
I don't have much experience with running applications on GPUs. So any help regarding this error will be highly appreciated.
Thanks,
Regards,
Aishwarya Sudhama