Dear All,

I am new to using Siesta. I have installed Siesta 3 to our system, and we
were trying some Test runs. I have acquired a sample run from a friend and
ran it to compare the results. I am using Transiesta for this test. My first
run, the Electrode run, is exactly the same, the .out file I use to flush
out the system output is exactly same with the sample.

However, when I run the second phase, that includes the scattering region, I
end up with a problem and my run won't converge. I went through the .out
files for my case and the sample. I recognized that after the following
entry, our runs start to differ.


New_DM. Step:     1
Initializing Density Matrix...
 *Warning: TSDM read from file has a different size
TRANSIESTA: No TS-DensityMatrix file found
TRANSIESTA: Initialization runs using diagon

The elec.TSHS file is there. I don't see any other file that has to be
present.

This Warning does not appear in the sample run .out file but only in my
file.

The problem is somehow more clarified once we look into the following entry:

siesta:    1   -17203.4037   -17390.3037   -17390.3037  1.3907 -6.9906
timer: Routine,Calls,Time,% = IterSCF        1     510.400  87.01
elaps: Routine,Calls,Wall,% = IterSCF        1      67.514  86.44
siesta:    2   -17199.2331   -17192.4978   -17192.4978  0.0664 -6.7745
siesta:    3   -17203.8611   -17195.1833   -17195.1833  0.1564 -7.2211
siesta:    4   -17421.4651   -17155.8094   -17155.8094  1.5248 -5.0611
siesta:    5   -47870.7899   -14290.8308   -14290.8308 17.9047-86.9153
siesta:    6  -587952.9491   207702.8292   207702.8292 26.1871********
siesta: WARNING: Qtot, Tr[D*S] =     448.000000     461.047491
siesta:    7 -1665888.2710   834710.1076   834710.1076 40.5776********
siesta: WARNING: Qtot, Tr[D*S] =     448.000000     461.048828
siesta:    8 -1666151.2923   834885.5125   834885.5125 40.5883********
siesta: WARNING: Qtot, Tr[D*S] =     448.000000     461.062600
siesta:    9 -1666175.8727   834959.1203   834959.1203 40.5280********
siesta: WARNING: Qtot, Tr[D*S] =     448.000000     461.045226
siesta:   10 -1660742.4149   831918.9038   831918.9038 39.9571********
siesta: WARNING: Qtot, Tr[D*S] =     448.000000     462.594965
siesta:   11 -1589022.4087   812183.5811   812183.5811 33.6616********
siesta: WARNING: Qtot, Tr[D*S] =     448.000000     467.029768
siesta:   12 -1838394.7537  1020260.0828  1020260.0828 39.9827********
siesta: WARNING: Qtot, Tr[D*S] =     448.000000     432.420602
siesta:   13 -2266090.6021  1167264.1830  1167264.1830 52.2963********
siesta: WARNING: Qtot, Tr[D*S] =     448.000000     433.329584
siesta:   14 -3645504.9501  2244783.9867  2244783.9867 52.4652********
siesta: WARNING: Qtot, Tr[D*S] =     448.000000     432.469152
siesta:   15 -4031151.9646  2546880.6182  2546880.6182 54.1435********
siesta: WARNING: Qtot, Tr[D*S] =     448.000000     432.459295
siesta:   16 -4040332.6834  2554005.0777  2554005.0777 54.1051********
siesta: WARNING: Qtot, Tr[D*S] =     448.000000     432.418588
siesta:   17 -4024189.5489  2540514.5799  2540514.5799 54.3196********
siesta: WARNING: Qtot, Tr[D*S] =     448.000000     432.532431
siesta:   18 -4017322.8332  2537964.4807  2537964.4807 53.8972********
siesta: WARNING: Qtot, Tr[D*S] =     448.000000     485.685367
siesta:   19 -7713161.0782  5544636.1008  5544636.1008 81.6685********
siesta: WARNING: Qtot, Tr[D*S] =     448.000000     481.639466
siesta:   20 -5567087.5426  3809614.6199  3809614.6199 69.6191********
siesta: WARNING: Qtot, Tr[D*S] =     448.000000     481.532817

There is clearly something odd going on here. Is my subroutines not
correctly compiled?

We have installed this program using GNU compilers, using the Intel MPI
compilers as backhand and the math libraries are from Intel MKL.

Thanks in advance,

Dr. Kemal Bagci

Responder a