rocsolver_ssyevd Interface Reference

rocsolver_ssyevd Interface Reference#

hipFORT API Reference: hipfort_rocsolver::rocsolver_ssyevd Interface Reference

The SYEVD functions compute the eigenvalues and optionally the eigenvectors of a real symmetric matrix A. More...

Public Member Functions
integer(c_int) function	rocsolver_ssyevd_ (handle, evect, uplo, n, a, lda, d, e, myinfo)

integer(kind(rocblas_status_success)) function	rocsolver_ssyevd_rank_0 (handle, evect, uplo, n, a, lda, d, e, myinfo)

integer(kind(rocblas_status_success)) function	rocsolver_ssyevd_rank_0_devptr (handle, evect, uplo, n, a, lda, d, e, myinfo)

integer(kind(rocblas_status_success)) function	rocsolver_ssyevd_rank_1 (handle, evect, uplo, n, a, lda, d, e, myinfo)

integer(kind(rocblas_status_success)) function	rocsolver_ssyevd_rank_1_devptr (handle, evect, uplo, n, a, lda, d, e, myinfo)

integer(kind(rocblas_status_success)) function	rocsolver_ssyevd_full_rank (handle, evect, uplo, n, a, lda, d, e, myinfo)

integer(kind(rocblas_status_success)) function	rocsolver_ssyevd_full_rank_devptr (handle, evect, uplo, n, a, lda, d, e, myinfo)

Detailed Description

The SYEVD functions compute the eigenvalues and optionally the eigenvectors of a real symmetric matrix A.

Note: The _64 interface accepts int64_t arguments, but the matrix dimensions n and lda must still fit within a 32-bit integer (less than 2^31). The internal tridiagonal reduction and back-transformation steps remain 32-bit, which bounds the supported size.

Note

A 2-stage tridiagonalization approach is available for SYEVD to improve performance for large matrices, n > 8000. Use rocsolver_set_alg_mode to enable it:

rocsolver_set_alg_mode( handle, rocsolver_function_sytrd, rocsolver_alg_mode_2stage );

Available modes are:

Mode	Description
rocsolver_alg_mode_1stage	Use 1-stage tridiagonalization algorithm (current default)
rocsolver_alg_mode_2stage	Use 2-stage tridiagonalization algorithm
rocsolver_alg_mode_auto	Automatically select the algorithm based on problem size

The eigenvalues are returned in ascending order. The eigenvectors are computed using a divide-and-conquer algorithm, depending on the value of evect. The computed eigenvectors are orthonormal.

Parameters

[in]	handle	- rocblas_handle.
[in]	evect	- `rocblas_evect`. Specifies whether the eigenvectors are to be computed. If evect is rocblas_evect_original, then the eigenvectors are computed. rocblas_evect_tridiagonal is not supported.
[in]	uplo	- rocblas_fill. Specifies whether the upper or lower part of the symmetric matrix A is stored. If uplo indicates lower (or upper), then the upper (or lower) part of A is not used.
[in]	n	- rocblas_int. n >= 0. Number of rows and columns of matrix A.
[in,out]	A	- pointer to type. Array on the GPU of dimension lda*n. On entry, the matrix A. On exit, the eigenvectors of A if they were computed and the algorithm converged. Otherwise, the contents of A are destroyed.
[in]	lda	- rocblas_int. lda >= n. Specifies the leading dimension of matrix A.
[out]	D	- pointer to type. Array on the GPU of dimension n. The eigenvalues of A in increasing order.
[out]	E	- pointer to type. Array on the GPU of dimension n. This array is used to work internally with the tridiagonal matrix T associated with A. On exit, if info > 0, it contains the unconverged off-diagonal elements of T (or properly speaking, a tridiagonal matrix equivalent to T). The diagonal elements of this matrix are in D. Those that converged correspond to a subset of the eigenvalues of A (not necessarily ordered).
[out]	myInfo	- pointer to a rocblas_int on the GPU. If info = 0, successful exit. If info = i > 0 and evect is rocblas_evect_none, the algorithm did not converge. i elements of E did not converge to zero. If info = i > 0 and evect is rocblas_evect_original, the algorithm failed to compute an eigenvalue in the submatrix from [i/(n+1), i/(n+1)] to [i%(n+1), i%(n+1)].

Member Function/Subroutine Documentation