<br>Dear Hui Wang,<br> But under such a case, should I using what kind of k-point? As following:<br><br>K_POINTS { tpiba | automatic | crystal | gamma }<br><br> gamma : use k = 0 ( do not read anything after this card )<br>
Note that a set of subroutines optimized for clculations at <br> the gamma point are used so that both memory and cpu requirements<br> are reduced<br> automatic: automatically generated uniform grid of k-points<br>
next card:<br> nk1, nk2, nk3, k1, k2, k3<br> generates ( nk1, nk2, nk3 ) mesh with ( k1, k2, k3 ) offset<br> nk1, nk2, nk3 as in Monkhorst-Pack grids<br> k1, k2, k3 must be 0 ( no offset ) or 1 ( grid displaced <br>
by half a grid step in the corresponding direction )<br> The mesh with offset may not work with tetrahedra.<br> crystal : read k-points in crystal coordinates<br> tpiba : read k-points in 2pi/a units ( default )<br>
next card:<br> nks<br> number of supplied special points<br> xk_x, xk_y, xk_z, wk<br> special points in the irreducible Brillouin Zone<br> of the lattice (with all symmetries) and weights<br>
If the symmetry is lower than the full symmetry <br> of the lattice, additional points with appropriate<br> weights are generated<br><br>===================================<br><br>And to save the time, what should I put in here? <br>
k point<br>1 1 4 <br>will be OK? Or either integer will be OK?<br><br>I am sorry for my bluntness.<br><br>Chengyu Yang<br>