rocsolver_sgeblttrs_npvt_interleaved_batched Interface Reference

rocsolver_sgeblttrs_npvt_interleaved_batched Interface Reference#

HIPFORT API Reference: hipfort_rocsolver::rocsolver_sgeblttrs_npvt_interleaved_batched Interface Reference

The GEBLTTRS_NPVT_INTERLEAVED_BATCHED functions solve a batch of system of linear equations given by block tridiagonal matrices in its factorized form (without partial pivoting). More...

Public Member Functions
integer(kind(rocblas_status_success)) function	rocsolver_sgeblttrs_npvt_interleaved_batched_ (handle, nb, nblocks, nrhs, a, inca, lda, stridea, b, incb, ldb, strideb, c, incc, ldc, stridec, x, incx, ldx, stridex, batch_count)

Detailed Description

The GEBLTTRS_NPVT_INTERLEAVED_BATCHED functions solve a batch of system of linear equations given by block tridiagonal matrices in its factorized form (without partial pivoting).

Each linear system has the form

\[ M_l X_l = \left[\begin{array}{ccccc} B_{l1} & C_{ll} & & & \\% A_{l1} & B_{ll} & C_{ll} & & \\% & \ddots & \ddots & \ddots & \\% & & A_{l(n-2)} & B_{l(n-1)} & C_{l(n-1)} \\% & & & A_{l(n-1)} & B_{ln} \end{array}\right]\left[\begin{array}{c} X_{l1}\\% X_{l2}\\% X_{l3}\\% \vdots\\% X_{ln} \end{array}\right]=\left[\begin{array}{c} R_{l1}\\% R_{l2}\\% R_{l3}\\% \vdots\\% R_{ln} \end{array}\right]=R_l \]

where matrix \(M_l\) has \(n = \) nblocks diagonal blocks of size nb, and the right-hand-side blocks \(R_{li}\) are general blocks of size nb -by-nrhs. The blocks of matrix \(M_l\) should be in the factorized form, as returned by GEBLTTRF_NPVT_INTERLEAVED_BATCHED.

Parameters

[in]	handle	- rocblas_handle.
[in]	nb	- rocblas_int. nb >= 0. The number of rows and columns of each block.
[in]	nblocks	- rocblas_int. nblocks >= 0. The number of blocks along the diagonal of each matrix in the batch.
[in]	nrhs	- rocblas_int. nrhs >= 0. The number of right hand sides, that is, the number of columns of blocks R_{li}.
[in]	A	- pointer to type. Array on the GPU (the size depends on the value of strideA). Contains the blocks A_{li}, as returned by GEBLTTRF_NPVT_INTERLEAVED_BATCHED.
[in]	inca	- rocblas_int. inca > 0. Stride from the start of one row of A_{li} to the next. The normal use cases are inca = 1 (equivalent to the strided batched case) or inca = batch_count (for an interleaved batched case).
[in]	lda	- rocblas_int. lda >= inca * nb. Specifies the leading dimension of blocks A_{li}, that is, the stride from the start of one column of A_{li} to the next.
[in]	strideA	- rocblas_stride. Stride from the start of one block A_{li} to the same block in the next batch instance A_{(l+1)i}. There is no restriction for the value of strideA. The normal use cases are strideA >= ldanb(nblocks-1) (equivalent to the strided batched case) or strideA = 1 (for an interleaved batched case).
[in]	B	- pointer to type. Array on the GPU (the size depends on the value of strideB). Contains the blocks B_{li}, as returned by GEBLTTRF_NPVT_INTERLEAVED_BATCHED.
[in]	incb	- rocblas_int. incb > 0. Stride from the start of one row of B_{li} to the next. The normal use cases are incb = 1 (equivalent to the strided batched case) or incb = batch_count (for an interleaved batched case).
[in]	ldb	- rocblas_int. ldb >= incb * nb. Specifies the leading dimension of blocks B_{li}, that is, the stride from the start of one column of B_{li} to the next.
[in]	strideB	- rocblas_stride. Stride from the start of one block B_{li} to the same block in the next batch instance B_{(l+1)i}. There is no restriction for the value of strideB. The normal use cases are strideB >= ldbnbnblocks (equivalent to the strided batched case) or strideB = 1 (for an interleaved batched case).
[in]	C	- pointer to type. Array on the GPU (the size depends on the value of strideC). Contains the blocks C_{li}, as returned by GEBLTTRF_NPVT_INTERLEAVED_BATCHED.
[in]	incc	- rocblas_int. incc > 0. Stride from the start of one row of C_{li} to the next. The normal use cases are incc = 1 (equivalent to the strided batched case) or incc = batch_count (for an interleaved batched case).
[in]	ldc	- rocblas_int. ldc >= incc * nb. Specifies the leading dimension of blocks C_{li}, that is, the stride from the start of one column of C_{li} to the next.
[in]	strideC	- rocblas_stride. Stride from the start of one block C_{li} to the same block in the next batch instance C_{(l+1)i}. There is no restriction for the value of strideC. The normal use cases are strideC >= ldcnb(nblocks-1) (equivalent to the strided batched case) or strideC = 1 (for an interleaved batched case).
[in,out]	X	- pointer to type. Array on the GPU (the size depends on the value of strideX). On entry, X contains the right-hand-side blocks R_{li}. It is overwritten by solution vectors X_{li} on exit.
[in]	incx	- rocblas_int. incx > 0. Stride from the start of one row of X_{li} to the next. The normal use cases are incx = 1 (equivalent to the strided batched case) or incx = batch_count (for an interleaved batched case).
[in]	ldx	- rocblas_int. ldx >= incx * nb. Specifies the leading dimension of blocks X_{li}, that is, the stride from the start of one column of X_{li} to the next.
[in]	strideX	- rocblas_stride. Stride from the start of one block X_{li} to the same block in the next batch instance X_{(l+1)i}. There is no restriction for the value of strideX. The normal use cases are strideX >= ldxnrhsnblocks (equivalent to the strided batched case) or strideX = 1 (for an interleaved batched case).
[in]	batch_count	- rocblas_int. batch_count >= 0. Number of matrices in the batch.

Member Function/Subroutine Documentation