Re: [SIESTA-L] convergence

I have energy doing up. What can be the major points to look in to change the situation? Is is possibly coming from the Broyden accelerator? siesta: iscf Eharris(eV) E_KS(eV) FreeEng(eV) dDmax Ef(eV) siesta:1 -128821.6727 -128821.6750 -128821.6819 0.0160 -2.6925 siesta:2

Re: [SIESTA-L] Bromine Pseudopotential

Hi David, The problem is not with the pseudopotential. You must use the block PS.lmax and set the max value of l of Br to 2, as in the siesta manual. Marcos > Dear siesta users, > > I have troubling in generate a Bromine pseudopotential that functions > in siesta. Anyone has a tested pseudopoten

Re: [SIESTA-L] Siesta-2.0 parallel version

El 16/06/2006, a las 0:39, Cherry Y. Yates escribió: It is not worth compiling it parallelly if you donot use a large number of k points (thousands). Its parallelism on gamma-point only is pretty poor. I found the best it can do is deceasing 30% time if running on a single node with two CPUs. If

[SIESTA-L] convergence

Hi, I have energy doing up. What can be the major points to look in to change the situation? Is is possibly coming from the Broyden accelerator? siesta: iscf Eharris(eV) E_KS(eV) FreeEng(eV) dDmax Ef(eV) siesta:1 -128821.6727 -128821.6750 -128821.6819 0.0160 -2.6925 siesta:

[SIESTA-L] Question about SIESTA-ART nouveau

Dear developers, I would like to ask whether the SIESTA-ART nouveau method for calculating diffusion pathway is implemented in SIESTA-2.0 ? This method was described in PHYSICAL REVIEW B 70, 205202 . Regards, H.P.

[SIESTA-L] MD in Siesta (fwd)

Gentlemen, may I ask once more about the way how to choose Nose parameter in MD ? Well, MD is not a frequent topic here, nor it appears frequently on the list of Siesta based papers, so a newbee is about to reinvent a wheel. Moreover, implementations of this option slightly vary from code to code s

Re: [SIESTA-L] Siesta-2.0 parallel version

It is not worth compiling it parallelly if you donot use a large number of k points (thousands). Its parallelism on gamma-point only is pretty poor. I found the best it can do is deceasing 30% time if running on a single node with two CPUs. If more nodes or more CPUs are used, essentially parallel

[SIESTA-L] Bromine Pseudopotential

Dear siesta users, I have troubling in generate a Bromine pseudopotential that functions in siesta. Anyone has a tested pseudopotential parameter for Bromine? I have tested the Bromine parameters of "atom_table.txt" that comes in /Contrib directory of siesta 2.0 program. The siesta 1.3.7 or siest