Page 1 of 1
ALGO=TDHF Error - TWOELECTRON4O_STORE_SCALA
Posted: Mon Aug 15, 2022 10:44 am
by chandan_kumarvishwakarma1
Dear users,
I am getting the same error i.e " internal error 4 in TWOELECTRON4O_STORE_SCALA: 2 1" with ALGO=TDHF.
I checked with smaller k-points and even with more computing nodes.
Please let me know about the same.
---
Chandan
Re: ALGO=TDHF Error - TWOELECTRON4O_STORE_SCALA
Posted: Tue Aug 16, 2022 2:08 pm
by alexey.tal
Hi,
Could you please provide all relevant input and output files for your calculation according to the forum
guidelines.
Re: ALGO=TDHF Error - TWOELECTRON4O_STORE_SCALA
Posted: Tue Aug 16, 2022 6:21 pm
by chandan_kumarvishwakarma1
Dear Alexey,
Here I am attaching a tar file including all the inputs,
Please refer, INCAR.DFT as a first step of mBSE calculation i.e. scf calculation and INCAR.mBSE as the INCAR for
the mBSE calculation.
I have checked my calculation using smaller kpoints 5x5x5 also but I am getting similar error.
Please find the attachment.
---
Chandan
Re: ALGO=TDHF Error - TWOELECTRON4O_STORE_SCALA
Posted: Wed Aug 17, 2022 8:45 am
by alexey.tal
If you reduce the k-points grid to 5x5x5, the BSE Hamiltonian rank according to your INCAR is 5^3*64*64=512000 and the required memory is rank^2. For double precision it is about 4TB. Are you sure that your nodes have enough memory?
You should try to reduce both the number of included states and k-points and see if your calculations finishes normally.