- Posts: 22
it can help to add cutoffs in the first EM for backward.py
- cneale
-
Topic Author
- Offline
- Junior Boarder
Less
More
2 years 9 months ago #9082
by cneale
it can help to add cutoffs in the first EM for backward.py was created by cneale
Hello,
I have a build that works with gmx 5.1.2 to convert from martini to charmm with the backward tool. However, It doesn't work with gmx2019.6. The error is on the first EM:
Fatal error:
Failed to allocate memory for 300124320 interactions, which is larger than the
maximum of 268435455. If your simulation is not using interaction cutoffs, you
should use them.
Adding the 0.9 nm cutoffs from the next step in intram.sh to the first EM solved this issue and the gmx2019.6 build now works. Perhaps 0.9 nm is too short for this first step, but I can't see why since bonded interactions are always going to be included.
The system is large (~1.5 million atoms). I suppose gromacs added some limit between 5.1.2 and 2019.6.
This is mostly just a report in case it helps anyone else.
I have a build that works with gmx 5.1.2 to convert from martini to charmm with the backward tool. However, It doesn't work with gmx2019.6. The error is on the first EM:
Fatal error:
Failed to allocate memory for 300124320 interactions, which is larger than the
maximum of 268435455. If your simulation is not using interaction cutoffs, you
should use them.
Adding the 0.9 nm cutoffs from the next step in intram.sh to the first EM solved this issue and the gmx2019.6 build now works. Perhaps 0.9 nm is too short for this first step, but I can't see why since bonded interactions are always going to be included.
The system is large (~1.5 million atoms). I suppose gromacs added some limit between 5.1.2 and 2019.6.
This is mostly just a report in case it helps anyone else.
Please Log in or Create an account to join the conversation.
Time to create page: 0.085 seconds