FATAL ERROR: CUDA error CudaComputeNonbonded::forceDoneCheck polled 1000000 times over 154.549621 s on Pe 4 (c36.chundoong device 0 pci 0:2:0)

From: 김민재 (kjh950429_at_gmail.com)
Date: Mon Apr 08 2019 - 23:00:38 CDT

Hi everyone,
I am currently trying to run an MD simulation (equilibriation to be exact,
the minimization went smoothly) with a protein-ligand system. I always get
the following error after a few thousand steps or so...
FATAL ERROR: CUDA error CudaComputeNonbonded::forceDoneCheck polled 1000000
times over 154.549621 s on Pe 4 (c36.chundoong device 0 pci 0:2:0)
[4] Stack Traceback:
  [4:0] [0x4fc92c]
  [4:1] [0xb13743]
  [4:2] [0xa8d14d]
  [4:3] [0x13a7a3d]
  [4:4] [0x139bab2]
  [4:5] [0x137e57e]
  [4:6] [0x137e059]
  [4:7] +0x7dc5 [0x7fe244a22dc5]
  [4:8] clone+0x6d [0x7fe24391221d]
[4] Stack Traceback:
  [4:0] [0x4fc92c]
  [4:1] [0xb13743]
  [4:2] [0xa8d14d]
  [4:3] [0x13a7a3d]
  [4:4] [0x139bab2]
  [4:5] [0x137e57e]
  [4:6] [0x137e059]
  [4:7] +0x7dc5 [0x7fe244a22dc5]
  [4:8] clone+0x6d [0x7fe24391221d]

I would deeply appreciate any suggestions on how to fix this issue.
Thanks

This archive was generated by hypermail 2.1.6 : Tue Dec 10 2019 - 23:20:41 CST