feature: parallel solve subspace diagonalization in dav_subspace #5549
Conversation
| - **Default**: 4 | ||
| - **Default**: 4 | ||
|
|
||
| ### diag_subspace_method |
There was a problem hiding this comment.
I would suggest using "diag_subspace" and delete "_method"
| const int* n, float* A, const int* ia, const int* ja, const int*desca, float* B, const int* ib, const int* jb, const int*descb, | ||
| const float* vl, const float* vu, const int* il, const int* iu, | ||
| const float* abstol, int* m, int* nz, float* w, const float*orfac, float* Z, const int* iz, const int* jz, const int*descz, | ||
| float* work, int* lwork, int*iwork, int*liwork, int* ifail, int*iclustr, float*gap, int* info); |
There was a problem hiding this comment.
at least leave a blank line between two functions
| const int diag_subspace_method, | ||
| const int block_size) | ||
| { | ||
| int myrank; |
There was a problem hiding this comment.
initialize the variable
|
|
||
| template <typename T> | ||
| void Diago_HS_para( | ||
| T* h, |
There was a problem hiding this comment.
I would suggset using clang-format before you pr a new code, here T is not aligned with the following variables
| work, lwork,rwork, lrwork, iwork, liwork, ifail, iclustr, gap, info); | ||
| } | ||
|
|
||
| // post processing for pdsygvx/pzhegvx/pdsygvx/pzhegvx |
There was a problem hiding this comment.
explain the functions of pdsygvx/pzhegvx/pdsygvx/pzhegvx
| typename GetTypeReal<T>::type *const ekb, | ||
| T *const wfc_2d) | ||
| { | ||
| int nprow, npcol, myprow, mypcol; |
There was a problem hiding this comment.
initialize one variable for a line
| int ndim_global = desc[2]; | ||
| const char jobz = 'V', range = 'I', uplo = 'U'; | ||
| const int itype = 1, il = 1, iu = nbands, one = 1; | ||
| int M = 0, NZ = 0, lwork = -1, lrwork = -1, liwork = -1, info = 0; |
There was a problem hiding this comment.
one line for a variable
| } | ||
|
|
||
| // template instantiation | ||
| template void pxxxgvx_diag(const int *const desc, |
There was a problem hiding this comment.
explain the functions of the four templates
|
|
|
I have a brief insight regarding this PR: This process involves a transformation of parallel strategy.
Can this transformation of parallel strategy be more general? If the H and S matrices of LCAO are solved by iterative method (like |
The parallel diagonization is not always more efficient, it is related to the system size, parallel cores, and efficiency of parallel communication, etc. I have done some tests in #5480. |
The transformation of different 2D block distribution can be realized easily by call the scalapack function |
| */ | ||
|
|
||
| template <typename T> | ||
| void pxxxgvx_diag(const int* const desc, |
There was a problem hiding this comment.
Does p stand for "parameter"?
There was a problem hiding this comment.
pxxxgvx is the wrapper of pdsygvx_/pzhegvx_/pdsygvx_/pzhegvx_ in scalapack
Diago_HS_para()to diago the H/S by genelpa or scalapack. The function will perform 2D block distribution of data, call elpa/scalapack for solution, and collect eigenvectors to rank0.pxxxgvx_diag()for calling scalapack to solve the generalized eigenvalue problem with double/complex/float/complex.diag_subspace_method, which support to set the subspace diagonalization method in dav_subspace to use elpa or scalapack to solve.Reminder
Linked Issue
Fix #5480
Unit Tests and/or Case Tests for my changes
What's changed?
Any changes of core modules? (ignore if not applicable)