DMRG is more sensitive to roundoff errors, which can be different with difference machines, compilers, compiler options, etc. Still the results should agree within the expected accuracies of the method ( number of states, convergence, etc.)
Best regards
Matthias Troyer
On 16 Jun 2016, at 20:47, Johannes Reim j.reim@tagen.tohoku.ac.jp wrote:
Dear all,
we used ALPS to determine the energy of a linear AF spin chain with DM on two different computer, both running Linux, which required us to build ALPS with the provided scripts. We have used the same parameter and model used, and compared the methods exact diagonalization and DMRG. While the diagonalization results are the same up to at least the 10th digit, the DMRG result on one computer differs from all others by about 0.3%. My question is, whether there exists an option to check the validity of the ALPS installation and which results are reliable.
Thank you very much and best wishes, Johannes
Comp-phys-alps-users Mailing List for the ALPS Project http://alps.comp-phys.org/
List info: https://lists.phys.ethz.ch//listinfo/comp-phys-alps-users Archive: https://lists.phys.ethz.ch//pipermail/comp-phys-alps-users
Unsubscribe by writing a mail to comp-phys-alps-users-leave@lists.phys.ethz.ch.