DIIS convergence acceleration

You can always use the default values. If the calculation exceeds the iteration limit before it reaches the convergence limit, then most likely there is a convergence failure. Simply increasing this limit is unlikely to help. The DIIS convergence accelerator may help in some cases. 

The DIIS convergence accelerator is available for all the SCF semiempirical methods. This accelerator may be helpful in curing convergence problems. It often reduces the number of iteration cycles required to reach convergence. However, it may be slower because it requires time to form a linear combination of the Fock matrices during the SCF calculation. The performance of the DIIS accelerator depends, in part, on the power of your computer. 

Choose the DIIS SCF convergence accelerator to potentially speed up SCF convergence. DIIS often reduces the number of iterations required to reach a convergence limit. However, it takes memory to store the Fock matrices from the previous iterations and this option may increase the computational time for individual iterations because the Fock matrix has to be calculated as a linear combination of the current Fock matrix and Fock matrices from previous iterations. 

Both schemes are also used as SCF convergence accelerators. The DIIS scheme is particularly efficient when used in conjunction with CPCM and IEFPCM schemes, in which the diagonal dominancy of T is less prominent than in DPCM. DIIS is very efficient from the point of view of CPU times, but it requires the storage of several sets of intermediate charges. DAMP is less efficient but requires the storage of two sets of intermediate charges only. 

The usual convergence acceleration/stabilisation tools may be employed in this orbital optimisation. For instance, we have implemented level shifting and DIIS [11]. 

When three-point interpolation fails to yield a convergent calculation, you can request a second accelerator for any SCFcalculation via the Semi-empirical Options dialog box and the Ab Initio Options dialog box. This alternative method. Direct Inversion in the Iterative Subspace (DIIS), was developed by Peter Pulay [P. Pulay, Chem. Phys. Lett., 73, 393 (1980) J. Comp. Chem., 3, 556(1982)]. DIIS relies on the fact that the eigenvectors of the density and Fock matrices are identical at self-consistency. At SCF convergence, the following condition exists... 

Having generated a sequence of optimization steps c , the Direct Inversion of the Iterative Subspace (DIIS) method (Csdszar and Pulay 1984 Hutter et al. 1994 Pulay 1980,1982) is designed to accelerate convergence by finding the best linear combination of stored c, vectors. 

In order to accelerate the SCF convergence or to overcome convergence problems, SlNDOl includes three different weighting procedures, a level shift procedure, a direct energy minimization procedure, and Pulay s direct inversion in iterative subspace (DIIS) procedure. ... 

The perturbation-based quasi-Newton scheme (13.4.10) is fairly robust, usually conveiging to six decimal places in the energy in 10-20 iterations, although as many as 50 iterations may be needed in difficult cases. As for Hartree-Fock theory in Section 10.6.2, the convergence may be improved significantly by application of the DIIS acceleration scheme. 

It should be realized that DUS by itself does not constitute an optimization algorithm. On its own, DUS does not produce an amplitude vector that cannot be written as a linear combination of those already generated. Rather, DUS provides an improved mechanism for utilizing the information contained in the quasi-Newton corrections (13.4.10). As we shall see in Section 13.4.4, the acceleration achieved by the DIIS procedure is often quite dramatic, significantly reducing the number of iterations needed for convergence. 

To illustrate the convergence rates of the perturbation-based quasi-Newton scheme with and without DIIS acceleration, we have in Table 13.2 listed the errors in the energy in each iteration of CCSD/cc-pVDZ calculations on the water molecule at the bond distances of / ref and 2/ ref- At R ei and 2Rref. the converged CCSD energies are —76.238116 and —75.929633 Eh, respectively (see Table 5.11). 

---