pycuda._driver.MemoryError: cuMemHostAlloc failed: out of memory

Hi all,

I always meet a err like this,but my RAM is 1T,and I think it is always enough.My OS is centos7,and my GPU is 3090.My CUDA is 11.2.When I do initial model,2d or some other jobs,it can give me that err.My boxsize is 320,and intial 4 model.Is there any suggestion for me?Thanks very much.I also change the ulimit -l as the suggestion in other similar topic.But it seems doesn’t work.I also add a 2T ssd to become the swap.Looking forward your suggestions.Thanks very much.
[CPU: 1.36 GB] Traceback (most recent call last):
File “cryosparc_worker/cryosparc_compute/”, line 84, in
File “cryosparc_worker/cryosparc_compute/jobs/abinit/”, line 304, in
File “cryosparc_worker/cryosparc_compute/engine/”, line 1149, in cryosparc_compute.engine.engine.process
File “cryosparc_worker/cryosparc_compute/engine/”, line 1150, in cryosparc_compute.engine.engine.process
File “cryosparc_worker/cryosparc_compute/engine/”, line 1089, in
File “cryosparc_worker/cryosparc_compute/engine/”, line 305, in cryosparc_compute.engine.engine.EngineThread.compute_resid_pow
File “cryosparc_worker/cryosparc_compute/engine/”, line 319, in cryosparc_compute.engine.cuda_core.EngineBaseThread.ensure_allocated
pycuda._driver.MemoryError: cuMemHostAlloc failed: out of memory