derek.holloway
7502bf64a4
CodeQL / Analyze (actions) (push) Canceled after 0s
CodeQL / Analyze (python) (push) Canceled after 0s
pre-commit / pre-commit (push) Canceled after 0s
Multi-Arch CI / setup (push) Canceled after 0s
Publish build_manylinux_x86_64 images / publish_build_manylinux_x86_64 (push) Canceled after 0s
Setup (Multi-Arch) / setup (push) Canceled after 0s
Multi-Arch CI / Linux::${{ fromJSON(needs.setup.outputs.linux_build_config || '{}').build_variant_label || 'skip' }} (push) Canceled after 0s
Publish TheRock Docker image / build-and-push-image (push) Canceled after 0s
Publish no_rocm_image_ubi10 image / publish_no_rocm_image_ubi10 (push) Canceled after 0s
Multi-Arch CI / Windows::${{ fromJSON(needs.setup.outputs.windows_build_config || '{}').build_variant_label || 'skip' }} (push) Canceled after 0s
Publish TheRock Docker image / build-and-push-image (push) Canceled after 0s
Multi-Arch CI / Manifest Diff (push) Canceled after 0s
TheRock Manifest Diff Report / Generate Manifest Diff Report (push) Canceled after 0s
Multi-Arch CI / CI Summary (push) Canceled after 0s
Publish no_rocm_image_ubuntu24_04 images / publish_no_rocm_image_ubuntu24_04 (push) Canceled after 0s
Publish no_rocm_image_ubuntu24_04_ocl_rt images / publish_no_rocm_image_ubuntu24_04_ocl_rt (push) Canceled after 0s
Publish TheRock Docker image / build-and-push-image (push) Canceled after 0s
Publish TheRock Docker image / build-and-push-image (push) Canceled after 0s
Publish no_rocm_image_ubuntu24_04 images / publish_no_rocm_image_ubuntu24_04_rocgdb (push) Canceled after 0s
Publish no_rocm_image_ubuntu24_04_rocgdb images / publish_no_rocm_image_ubuntu24_04_rocgdb (push) Canceled after 0s
Publish no_rocm_image_ubuntu24_04 images / publish_no_rocm_image_ubuntu24_04_ocl_rt (push) Canceled after 0s
Publish TheRock Docker image / build-and-push-image (push) Canceled after 0s
Unit Tests / Unit Tests :: ubuntu-24.04 (push) Canceled after 0s
Unit Tests / Unit Tests :: windows-2022 (push) Canceled after 0s
Publish no_rocm_image_ubuntu24_04 images / publish_no_rocm_image_ubuntu24_04_media (push) Canceled after 0s
Unit Tests / Unit Tests Summary (push) Canceled after 0s
Publish no_rocm_image_ubuntu24_04 images / publish_no_rocm_image_ubuntu24_04_openmpi (push) Canceled after 0s
387 lines
18 KiB
C
387 lines
18 KiB
C
/*! \file */
|
|
/* ************************************************************************
|
|
* Copyright (C) 2025-2026 Advanced Micro Devices, Inc. All rights Reserved.
|
|
*
|
|
* Permission is hereby granted, free of charge, to any person obtaining a copy
|
|
* of this software and associated documentation files (the "Software"), to deal
|
|
* in the Software without restriction, including without limitation the rights
|
|
* to use, copy, modify, merge, publish, distribute, sublicense, and/or sell
|
|
* copies of the Software, and to permit persons to whom the Software is
|
|
* furnished to do so, subject to the following conditions:
|
|
*
|
|
* The above copyright notice and this permission notice shall be included in
|
|
* all copies or substantial portions of the Software.
|
|
*
|
|
* THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
|
|
* IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
|
|
* FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE
|
|
* AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER
|
|
* LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM,
|
|
* OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN
|
|
* THE SOFTWARE.
|
|
*
|
|
* ************************************************************************ */
|
|
#ifndef HIPSPARSE_SPMM_H
|
|
#define HIPSPARSE_SPMM_H
|
|
|
|
#ifdef __cplusplus
|
|
extern "C" {
|
|
#endif
|
|
|
|
/*! \ingroup generic_module
|
|
* \details
|
|
* \p hipsparseSpMM_bufferSize computes the required user-allocated buffer size needed when computing the
|
|
* sparse matrix multiplication with a dense matrix:
|
|
* \f[
|
|
* C := \alpha \cdot op(A) \cdot op(B) + \beta \cdot C,
|
|
* \f]
|
|
* where \f$op(A)\f$ is a sparse \f$m \times k\f$ matrix in CSR, CSC, COO, BSR, or Blocked ELL storage format,
|
|
* \f$B\f$ is a dense matrix of size \f$k \times n\f$, and \f$C\f$ is a dense matrix of size \f$m \times n\f$.
|
|
*
|
|
* \p hipsparseSpMM_bufferSize supports multiple combinations of data types and compute types. See \ref hipsparseSpMM
|
|
* for a complete listing of all the data type and compute type combinations available.
|
|
*
|
|
* @param[in]
|
|
* handle handle to the hipSPARSE library context queue.
|
|
* @param[in]
|
|
* opA matrix operation type.
|
|
* @param[in]
|
|
* opB matrix operation type.
|
|
* @param[in]
|
|
* alpha scalar \f$\alpha\f$.
|
|
* @param[in]
|
|
* matA sparse matrix descriptor.
|
|
* @param[in]
|
|
* matB dense matrix descriptor.
|
|
* @param[in]
|
|
* beta scalar \f$\beta\f$.
|
|
* @param[in]
|
|
* matC dense matrix descriptor.
|
|
* @param[in]
|
|
* computeType floating point precision for the SpMM computation.
|
|
* @param[in]
|
|
* alg SpMM algorithm for the SpMM computation.
|
|
* @param[out]
|
|
* pBufferSizeInBytes number of bytes of the temporary storage buffer.
|
|
*
|
|
* \retval HIPSPARSE_STATUS_SUCCESS the operation completed successfully.
|
|
* \retval HIPSPARSE_STATUS_NOT_INITIALIZED \p handle is not initialized.
|
|
* \retval HIPSPARSE_STATUS_INVALID_VALUE \p handle, \p alpha, \p matA, \p matB, \p matC, \p beta, or
|
|
* \p pBufferSizeInBytes is nullptr, or \p opA or \p opB is invalid.
|
|
* \retval HIPSPARSE_STATUS_NOT_SUPPORTED \p opA, \p opB, \p computeType, or \p alg is currently not supported.
|
|
*/
|
|
#if(!defined(CUDART_VERSION) || CUDART_VERSION >= 12000)
|
|
HIPSPARSE_EXPORT
|
|
hipsparseStatus_t hipsparseSpMM_bufferSize(hipsparseHandle_t handle,
|
|
hipsparseOperation_t opA,
|
|
hipsparseOperation_t opB,
|
|
const void* alpha,
|
|
hipsparseConstSpMatDescr_t matA,
|
|
hipsparseConstDnMatDescr_t matB,
|
|
const void* beta,
|
|
const hipsparseDnMatDescr_t matC,
|
|
hipDataType computeType,
|
|
hipsparseSpMMAlg_t alg,
|
|
size_t* pBufferSizeInBytes);
|
|
#elif(CUDART_VERSION >= 10010)
|
|
HIPSPARSE_EXPORT
|
|
hipsparseStatus_t hipsparseSpMM_bufferSize(hipsparseHandle_t handle,
|
|
hipsparseOperation_t opA,
|
|
hipsparseOperation_t opB,
|
|
const void* alpha,
|
|
const hipsparseSpMatDescr_t matA,
|
|
const hipsparseDnMatDescr_t matB,
|
|
const void* beta,
|
|
const hipsparseDnMatDescr_t matC,
|
|
hipDataType computeType,
|
|
hipsparseSpMMAlg_t alg,
|
|
size_t* pBufferSizeInBytes);
|
|
#endif
|
|
|
|
/*! \ingroup generic_module
|
|
* \details
|
|
* \p hipsparseSpMM_preprocess performs the required preprocessing used when computing the
|
|
* sparse matrix multiplication with a dense matrix:
|
|
* \f[
|
|
* C := \alpha \cdot op(A) \cdot op(B) + \beta \cdot C,
|
|
* \f]
|
|
* where \f$op(A)\f$ is a sparse \f$m \times k\f$ matrix in CSR, CSC, COO, BSR, or Blocked ELL storage format,
|
|
* \f$B\f$ is a dense matrix of size \f$k \times n\f$, and \f$C\f$ is a dense matrix of size \f$m \times n\f$.
|
|
*
|
|
* \p hipsparseSpMM_preprocess supports multiple combinations of data types and compute types. See \ref hipsparseSpMM for a complete
|
|
* listing of all the data type and compute type combinations available.
|
|
*
|
|
* @param[in]
|
|
* handle handle to the hipSPARSE library context queue.
|
|
* @param[in]
|
|
* opA matrix operation type.
|
|
* @param[in]
|
|
* opB matrix operation type.
|
|
* @param[in]
|
|
* alpha scalar \f$\alpha\f$.
|
|
* @param[in]
|
|
* matA matrix descriptor.
|
|
* @param[in]
|
|
* matB matrix descriptor.
|
|
* @param[in]
|
|
* beta scalar \f$\beta\f$.
|
|
* @param[in]
|
|
* matC matrix descriptor.
|
|
* @param[in]
|
|
* computeType floating point precision for the SpMM computation.
|
|
* @param[in]
|
|
* alg SpMM algorithm for the SpMM computation.
|
|
* @param[out]
|
|
* externalBuffer temporary storage buffer allocated by the user.
|
|
*
|
|
* \retval HIPSPARSE_STATUS_SUCCESS the operation completed successfully.
|
|
* \retval HIPSPARSE_STATUS_INVALID_VALUE \p handle, \p alpha, \p matA, \p matB, \p matC, \p beta, or
|
|
* \p externalBuffer pointer is invalid.
|
|
* \retval HIPSPARSE_STATUS_NOT_SUPPORTED \p opA, \p opB, \p computeType, or \p alg is
|
|
* currently not supported.
|
|
*/
|
|
#if(!defined(CUDART_VERSION) || CUDART_VERSION >= 12000)
|
|
HIPSPARSE_EXPORT
|
|
hipsparseStatus_t hipsparseSpMM_preprocess(hipsparseHandle_t handle,
|
|
hipsparseOperation_t opA,
|
|
hipsparseOperation_t opB,
|
|
const void* alpha,
|
|
hipsparseConstSpMatDescr_t matA,
|
|
hipsparseConstDnMatDescr_t matB,
|
|
const void* beta,
|
|
const hipsparseDnMatDescr_t matC,
|
|
hipDataType computeType,
|
|
hipsparseSpMMAlg_t alg,
|
|
void* externalBuffer);
|
|
#elif(CUDART_VERSION >= 11021)
|
|
HIPSPARSE_EXPORT
|
|
hipsparseStatus_t hipsparseSpMM_preprocess(hipsparseHandle_t handle,
|
|
hipsparseOperation_t opA,
|
|
hipsparseOperation_t opB,
|
|
const void* alpha,
|
|
const hipsparseSpMatDescr_t matA,
|
|
const hipsparseDnMatDescr_t matB,
|
|
const void* beta,
|
|
const hipsparseDnMatDescr_t matC,
|
|
hipDataType computeType,
|
|
hipsparseSpMMAlg_t alg,
|
|
void* externalBuffer);
|
|
#endif
|
|
|
|
/*! \ingroup generic_module
|
|
* \brief Compute the sparse matrix multiplication with a dense matrix.
|
|
*
|
|
* \details
|
|
* \p hipsparseSpMM multiplies the scalar \f$\alpha\f$ with a sparse \f$m \times k\f$ matrix \f$op(A)\f$,
|
|
* defined in CSR, CSC, COO, BSR, or Blocked ELL storage format, and the dense \f$k \times n\f$ matrix \f$op(B)\f$
|
|
* and adds the result to the dense \f$m \times n\f$ matrix \f$C\f$ that is multiplied by the scalar
|
|
* \f$\beta\f$, such that
|
|
* \f[
|
|
* C := \alpha \cdot op(A) \cdot op(B) + \beta \cdot C,
|
|
* \f]
|
|
* with
|
|
* \f[
|
|
* op(A) = \left\{
|
|
* \begin{array}{ll}
|
|
* A, & \text{if trans_A == HIPSPARSE_OPERATION_NON_TRANSPOSE} \\
|
|
* A^T, & \text{if trans_A == HIPSPARSE_OPERATION_TRANSPOSE} \\
|
|
* A^H, & \text{if trans_A == HIPSPARSE_OPERATION_CONJUGATE_TRANSPOSE}
|
|
* \end{array}
|
|
* \right.
|
|
* \f]
|
|
* and
|
|
* \f[
|
|
* op(B) = \left\{
|
|
* \begin{array}{ll}
|
|
* B, & \text{if trans_B == HIPSPARSE_OPERATION_NON_TRANSPOSE} \\
|
|
* B^T, & \text{if trans_B == HIPSPARSE_OPERATION_TRANSPOSE} \\
|
|
* B^H, & \text{if trans_B == HIPSPARSE_OPERATION_CONJUGATE_TRANSPOSE}
|
|
* \end{array}
|
|
* \right.
|
|
* \f]
|
|
* Both \f$B\f$ and \f$C\f$ can be in row or column order.
|
|
*
|
|
* \p hipsparseSpMM requires three stages to complete. First, the user calls \ref hipsparseSpMM_bufferSize to determine
|
|
* the size of the required temporary storage buffer. Next, the user allocates this buffer and calls
|
|
* \ref hipsparseSpMM_preprocess, which will perform analysis on the sparse matrix \f$op(A)\f$. Finally, the user calls
|
|
* \p hipsparseSpMM to perform the actual computation. The buffer size and preprocess routines only need to be called once for a given
|
|
* sparse matrix \f$op(A)\f$, while the computation routine can be repeatedly used with different \f$B\f$ and \f$C\f$ matrices.
|
|
* After all calls to \p hipsparseSpMM are complete, the temporary buffer can be deallocated.
|
|
*
|
|
* As noted above, both \f$B\f$ and \f$C\f$ can be in row or column order (this includes mixing the order so that \f$B\f$ is
|
|
* row order and \f$C\f$ is column order and vice versa). For best performance, use row order for both \f$B\f$ and \f$C\f$ because
|
|
* this provides the best memory access.
|
|
*
|
|
* \p hipsparseSpMM supports multiple different algorithms. These algorithms have different trade offs depending on the sparsity
|
|
* pattern of the matrix, whether or not the results need to be deterministic, and how many times the sparse-matrix product will
|
|
* be performed.
|
|
*
|
|
* <table>
|
|
* <caption id="spmm_csr_algorithms">CSR/CSC Algorithms</caption>
|
|
* <tr><th>CSR/CSC Algorithms
|
|
* <tr><td>HIPSPARSE_SPMM_CSR_ALG1</td>
|
|
* <tr><td>HIPSPARSE_SPMM_CSR_ALG2</td>
|
|
* <tr><td>HIPSPARSE_SPMM_CSR_ALG3</td>
|
|
* </table>
|
|
*
|
|
* <table>
|
|
* <caption id="spmm_coo_algorithms">COO Algorithms</caption>
|
|
* <tr><th>COO Algorithms
|
|
* <tr><td>HIPSPARSE_SPMM_COO_ALG1</td>
|
|
* <tr><td>HIPSPARSE_SPMM_COO_ALG2</td>
|
|
* <tr><td>HIPSPARSE_SPMM_COO_ALG3</td>
|
|
* <tr><td>HIPSPARSE_SPMM_COO_ALG4</td>
|
|
* </table>
|
|
*
|
|
* <table>
|
|
* <caption id="spmm_bell_algorithms">Blocked-ELL Algorithms</caption>
|
|
* <tr><th>ELL Algorithms
|
|
* <tr><td>HIPSPARSE_SPMM_BLOCKED_ELL_ALG1</td>
|
|
* </table>
|
|
*
|
|
* <table>
|
|
* <caption id="spmm_bsr_algorithms">BSR Algorithms</caption>
|
|
* <tr><th>BSR Algorithms
|
|
* <tr><td>HIPSPARSE_SPMM_BSR_ALG1</td>
|
|
* </table>
|
|
*
|
|
* Users can also pass \ref HIPSPARSE_SPMM_ALG_DEFAULT, which will automatically select from the algorithms listed above
|
|
* based on the sparse matrix format.
|
|
*
|
|
* When A is transposed, \p hipsparseSpMM will revert to using \ref HIPSPARSE_SPMM_CSR_ALG2
|
|
* for CSR format and \ref HIPSPARSE_SPMM_COO_ALG1 for COO format, regardless of the algorithm selected.
|
|
*
|
|
* \p hipsparseSpMM supports multiple combinations of data types and compute types. The tables below indicate the currently
|
|
* supported different data types that can be used for for the sparse matrix \f$op(A)\f$, the dense matrices \f$op(B)\f$ and
|
|
* \f$C\f$, and the compute type for \f$\alpha\f$ and \f$\beta\f$. The advantage of using different data types is to save on
|
|
* memory bandwidth and storage when the user application supports it, while performing the actual computation in a higher precision.
|
|
*
|
|
* \par Uniform Precisions:
|
|
* <table>
|
|
* <caption id="spmm_uniform">Uniform Precisions</caption>
|
|
* <tr><th>A / B / C / compute_type
|
|
* <tr><td>HIP_R_32F
|
|
* <tr><td>HIP_R_64F
|
|
* <tr><td>HIP_C_32F
|
|
* <tr><td>HIP_C_64F
|
|
* </table>
|
|
*
|
|
* \par Mixed Precisions:
|
|
* <table>
|
|
* <caption id="spmm_mixed">Mixed Precisions</caption>
|
|
* <tr><th>A / B <th>C <th>compute_type
|
|
* <tr><td>HIP_R_8I <td>HIP_R_32I <td>HIP_R_32I
|
|
* <tr><td>HIP_R_8I <td>HIP_R_32F <td>HIP_R_32F
|
|
* <tr><td>HIP_R_16F <td>HIP_R_32F <td>HIP_R_32F
|
|
* <tr><td>HIP_R_16F <td>HIP_R_16F <td>HIP_R_32F
|
|
* <tr><td>HIP_R_16BF <td>HIP_R_32F <td>HIP_R_32F
|
|
* <tr><td>HIP_R_16BF <td>HIP_R_16BF <td>HIP_R_32F
|
|
* </table>
|
|
*
|
|
* \p hipsparseSpMM supports \ref HIPSPARSE_INDEX_32I and \ref HIPSPARSE_INDEX_64I index precisions
|
|
* for storing the row pointer and column indices arrays of the sparse matrices.
|
|
*
|
|
* \p hipsparseSpMM also supports batched computation for CSR, CSC, COO and Blocked ELL matrices. There are three supported batch modes:
|
|
* \f[
|
|
* C_i = A \times B_i \\
|
|
* C_i = A_i \times B \\
|
|
* C_i = A_i \times B_i
|
|
* \f]
|
|
*
|
|
* The batch mode is determined by the batch count and stride passed for each matrix. For example,
|
|
* to use the first batch mode (\f$C_i = A \times B_i\f$) with 100 batches for non-transposed \f$A\f$,
|
|
* \f$B\f$, and \f$C\f$, pass the following:
|
|
* \f[
|
|
* batchCountA=1 \\
|
|
* batchCountB=100 \\
|
|
* batchCountC=100 \\
|
|
* offsetsBatchStrideA=0 \\
|
|
* columnsValuesBatchStrideA=0 \\
|
|
* batchStrideB=k*n \\
|
|
* batchStrideC=m*n
|
|
* \f]
|
|
* To use the second batch mode (\f$C_i = A_i \times B\f$), use these settings:
|
|
* \f[
|
|
* batchCountA=100 \\
|
|
* batchCountB=1 \\
|
|
* batchCountC=100 \\
|
|
* offsetsBatchStrideA=m+1 \\
|
|
* columnsValuesBatchStrideA=nnz \\
|
|
* batchStrideB=0 \\
|
|
* batchStrideC=m*n
|
|
* \f]
|
|
* And to use the third batch mode (\f$C_i = A_i \times B_i\f$), pass these values:
|
|
* \f[
|
|
* batchCountA=100 \\
|
|
* batchCountB=100 \\
|
|
* batchCountC=100 \\
|
|
* offsetsBatchStrideA=m+1 \\
|
|
* columnsValuesBatchStrideA=nnz \\
|
|
* batchStrideB=k*n \\
|
|
* batchStrideC=m*n
|
|
* \f]
|
|
* For more information, see the examples below.
|
|
*
|
|
* @param[in]
|
|
* handle handle to the hipSPARSE library context queue.
|
|
* @param[in]
|
|
* opA matrix operation type.
|
|
* @param[in]
|
|
* opB matrix operation type.
|
|
* @param[in]
|
|
* alpha scalar \f$\alpha\f$.
|
|
* @param[in]
|
|
* matA matrix descriptor.
|
|
* @param[in]
|
|
* matB matrix descriptor.
|
|
* @param[in]
|
|
* beta scalar \f$\beta\f$.
|
|
* @param[in]
|
|
* matC matrix descriptor.
|
|
* @param[in]
|
|
* computeType floating point precision for the SpMM computation.
|
|
* @param[in]
|
|
* alg SpMM algorithm for the SpMM computation.
|
|
* @param[out]
|
|
* externalBuffer temporary storage buffer allocated by the user.
|
|
*
|
|
* \retval HIPSPARSE_STATUS_SUCCESS the operation completed successfully.
|
|
* \retval HIPSPARSE_STATUS_INVALID_VALUE \p handle, \p alpha, \p matA, \p matB, \p matC, \p beta, or
|
|
* \p externalBuffer pointer is invalid.
|
|
* \retval HIPSPARSE_STATUS_NOT_SUPPORTED \p opA, \p opB, \p computeType, or \p alg is
|
|
* currently not supported.
|
|
*/
|
|
#if(!defined(CUDART_VERSION) || CUDART_VERSION >= 12000)
|
|
HIPSPARSE_EXPORT
|
|
hipsparseStatus_t hipsparseSpMM(hipsparseHandle_t handle,
|
|
hipsparseOperation_t opA,
|
|
hipsparseOperation_t opB,
|
|
const void* alpha,
|
|
hipsparseConstSpMatDescr_t matA,
|
|
hipsparseConstDnMatDescr_t matB,
|
|
const void* beta,
|
|
const hipsparseDnMatDescr_t matC,
|
|
hipDataType computeType,
|
|
hipsparseSpMMAlg_t alg,
|
|
void* externalBuffer);
|
|
#elif(CUDART_VERSION >= 10010)
|
|
HIPSPARSE_EXPORT
|
|
hipsparseStatus_t hipsparseSpMM(hipsparseHandle_t handle,
|
|
hipsparseOperation_t opA,
|
|
hipsparseOperation_t opB,
|
|
const void* alpha,
|
|
const hipsparseSpMatDescr_t matA,
|
|
const hipsparseDnMatDescr_t matB,
|
|
const void* beta,
|
|
const hipsparseDnMatDescr_t matC,
|
|
hipDataType computeType,
|
|
hipsparseSpMMAlg_t alg,
|
|
void* externalBuffer);
|
|
#endif
|
|
|
|
#ifdef __cplusplus
|
|
}
|
|
#endif
|
|
|
|
#endif /* HIPSPARSE_SPMM_H */
|