The following parameters (collected in file param.inc_r or param.inc_c) are used:
HB Hierarchical blocking factor for the modified LAPACK matrix routines.
HB should be optimized for your hardware (machine type and cache
size). (Recommended: HB=32 for most cache based workstations,
HB=255 on NEC-SX4 vector computer).
HB=1 leads to use of the ``conventional'' matrix
diagonalization routines
(might still be the fastest when no optimized BLAS
routines are available).
KMAX1 the highest h,k,l Fourier components of the potential
KMAX2
KMAX3
LMAX highest l+1 in basis function inside sphere (consistent
with input in case.in1)
LMMX number of LM terms in potential (should be at least
NCOM-1)
LOMAX highest l for local orbital basis (consistent with input
in case.in1)
NATO number of inequivalent atoms
NDIF total number of atoms per unit cell
NGAU number of Gaunt coefficients for the non-spherical contributions
to the matrix elements
NKPT total number of k-points in irreducible wedge of Brillouin
zone
NMAT size of matrix (defines size of program!)
NMATIT should be the same as NMAT, when using ``iterative
diagonalization''. For ``full diagonalization''
memory size can be reduced by setting NMATIT=1
NRAD number of radial mesh points
NSLMAX highest l+1 in basis functions for non-muffin-tin matrix
elements (consistent with input in case.in1)
NSYM order of point group
NUME maximum number of energy eigenvalues per k-point
NUMEIT should be the same as NUME, when using ``iterative
diagonalization''. For ``full diagonalization''
memory size can be reduced by setting NUMEIT=1
NVEC1 defines the largest triple of integers which define
reciprocal
NVEC2 K-vectors when multiplied with the reciprocal Bravais
matrix
NVEC3
NWAV max. number of Fourier coefficients of nonspherical potential