If you have problems during the execution of MRCC, please attach the output with an adequate description of your case as well as the followings:
  • the way mrcc was invoked
  • the way build.mrcc was invoked
  • the output of build.mrcc
  • compiler version (for example: ifort -V, gfortran -v)
  • blas/lapack versions
  • as well as gcc and glibc versions

This information really helps us during troubleshooting :)

Parallelization of LNO-CCSD(T)

  • benshi97
  • Topic Author
  • Offline
  • New Member
  • New Member
More
2 years 8 months ago #1135 by benshi97
Parallelization of LNO-CCSD(T) was created by benshi97
Dear MRCC forum,

From reading the papers for LNO-CCSD(T), it is clear that there is OpenMP parallelization for the LNO-CCSD(T) code within a single node, but is it also able to utilize multi-node MPI parallelization? In the section describing the parallelization of MRCC in the JCP review paper, all tests and descriptions seemed to be for DF-CCSD(T), with no mention of LNO-CCSD(T) so it is not clear to me.

Best wishes,
Benjamin

Please Log in or Create an account to join the conversation.

  • benshi97
  • Topic Author
  • Offline
  • New Member
  • New Member
More
2 years 8 months ago #1138 by benshi97
Replied by benshi97 on topic Parallelization of LNO-CCSD(T)
Nevermind, after compiling in MPI and testing LNO-CCSD(T) calculations, MRCC says that it can't run local correlation methods across nodes.

Please Log in or Create an account to join the conversation.

  • nagypeter
  • Offline
  • Premium Member
  • Premium Member
  • MRCC developer
More
2 years 8 months ago #1139 by nagypeter
Replied by nagypeter on topic Parallelization of LNO-CCSD(T)
Dear Benjamin,

thanks for the interest.
Indeed, the MPI parallel LNO-CCSD(T) code is still under development and not yet released.

With the current release you can do some complicated tricks to exploit the MPI-OpenMP feature of the DF-CCSD(T) code within an LNO-CCSD(T) calculation.
However, I would not suggest such complicated workarounds as you just started using MRCC.

I think you should first collect some experience using LNO-CCSD(T) with your system of interest, as it might turn out the you could manage without the MPI parallelization.
To illustrate the capabilities some large scale applications (with OpenMP only execution) are collected here:
pubs.acs.org/doi/abs/10.1021/acs.jctc.9b00511

Best wishes,
Peter

Please Log in or Create an account to join the conversation.

Time to create page: 0.040 seconds
Powered by Kunena Forum