Files
ceres-solver/internal/ceres/linear_solver.cc
T
Joydeep Biswas 829089053e CUDA CGNR, Part 4: CudaCgnrSolver
* Added CudaCgnrSolver, a new CUDA-accelerated CGNR.
* To use CudaCgnrSolver, the user must select CGNR as the linear_solver
  and CUDA_SPARSE as the sparse_linear_algebra_library.
* Updated ConjugateGradientSolver to work with an array of pointers to
  scratch to support CudaVectors as scratch.
* Moved CUDA initialization to run in Solver::Solve as needed.

Some performance comparisons on an Ubuntu 20.04 desktop with an
Intel i9-9940X CPU @ 3.30GHz, and an nVidia Quadro RTX 6000,
all configurations run with 24 threads, and 10 iterations.

=================================================
CGNR + CUDA_SPARSE + IDENTITY Preconditioner
problem-1778-993923-pre.txt
=================================================
Cost:
Initial                          2.563973e+08
Final                            1.724755e+06
Change                           2.546725e+08

Minimizer iterations                       11
Successful steps                            7
Unsuccessful steps                          4

Time (in seconds):
Preprocessor                         4.020158

  Residual only evaluation           1.567092 (10)
  Jacobian & residual evaluation     7.847130 (7)
  Linear solver                     31.688898 (10)
Minimizer                           46.834987

Postprocessor                        0.353974
Total                               51.209120

=================================================
SPARSE_SCHUR (CPU) + SUITE_SPARSE + AMD
problem-1778-993923-pre.txt
=================================================
Cost:
Initial                          2.563973e+08
Final                            1.651617e+06
Change                           2.547457e+08

Minimizer iterations                       11
Successful steps                           11
Unsuccessful steps                          0

Time (in seconds):
Preprocessor                        35.812003

  Residual only evaluation           1.658980 (10)
  Jacobian & residual evaluation    12.218799 (11)
  Linear solver                     76.409992 (10)
Minimizer                           98.809773

Postprocessor                        0.372712
Total                              134.994489

=================================================
ITERATIVE_SCHUR (CPU) + JACOBI Preconditioner
problem-1778-993923-pre.txt
=================================================
Cost:
Initial                          2.563973e+08
Final                            1.684447e+06
Change                           2.547128e+08

Minimizer iterations                       11
Successful steps                            8
Unsuccessful steps                          3

Time (in seconds):
Preprocessor                        15.331614

  Residual only evaluation           1.606114 (10)
  Jacobian & residual evaluation     8.502166 (8)
  Linear solver                    351.910080 (10)
Minimizer                          368.797327

Postprocessor                        0.363536
Total                              384.492478

=================================================
CGNR + CUDA_SPARSE + IDENTITY Preconditioner
problem-13682-4456117-pre.txt
=================================================
Cost:
Initial                          1.126372e+09
Final                            2.269329e+07
Change                           1.103678e+09

Minimizer iterations                       11
Successful steps                            7
Unsuccessful steps                          4

Time (in seconds):
Preprocessor                        19.140087

  Residual only evaluation           8.721920 (10)
  Jacobian & residual evaluation    41.955923 (7)
  Linear solver                    214.121861 (10)
Minimizer                          296.636890

Postprocessor                        1.971827
Total                              317.748804

Change-Id: I3a09f31aa6903f661e91f595afd39d427583e856
2022-08-15 23:44:34 -05:00

130 lines
4.3 KiB
C++

// Ceres Solver - A fast non-linear least squares minimizer
// Copyright 2015 Google Inc. All rights reserved.
// http://ceres-solver.org/
//
// Redistribution and use in source and binary forms, with or without
// modification, are permitted provided that the following conditions are met:
//
// * Redistributions of source code must retain the above copyright notice,
// this list of conditions and the following disclaimer.
// * Redistributions in binary form must reproduce the above copyright notice,
// this list of conditions and the following disclaimer in the documentation
// and/or other materials provided with the distribution.
// * Neither the name of Google Inc. nor the names of its contributors may be
// used to endorse or promote products derived from this software without
// specific prior written permission.
//
// THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS "AS IS"
// AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE
// IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE
// ARE DISCLAIMED. IN NO EVENT SHALL THE COPYRIGHT OWNER OR CONTRIBUTORS BE
// LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR
// CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF
// SUBSTITUTE GOODS OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS
// INTERRUPTION) HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN
// CONTRACT, STRICT LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE)
// ARISING IN ANY WAY OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE
// POSSIBILITY OF SUCH DAMAGE.
//
// Author: sameeragarwal@google.com (Sameer Agarwal)
#include "ceres/linear_solver.h"
#include <memory>
#include "ceres/cgnr_solver.h"
#include "ceres/dense_normal_cholesky_solver.h"
#include "ceres/dense_qr_solver.h"
#include "ceres/dynamic_sparse_normal_cholesky_solver.h"
#include "ceres/internal/config.h"
#include "ceres/iterative_schur_complement_solver.h"
#include "ceres/schur_complement_solver.h"
#include "ceres/sparse_normal_cholesky_solver.h"
#include "ceres/types.h"
#include "glog/logging.h"
namespace ceres::internal {
LinearSolver::~LinearSolver() = default;
LinearSolverType LinearSolver::LinearSolverForZeroEBlocks(
LinearSolverType linear_solver_type) {
if (!IsSchurType(linear_solver_type)) {
return linear_solver_type;
}
if (linear_solver_type == SPARSE_SCHUR) {
return SPARSE_NORMAL_CHOLESKY;
}
if (linear_solver_type == DENSE_SCHUR) {
// TODO(sameeragarwal): This is probably not a great choice.
// Ideally, we should have a DENSE_NORMAL_CHOLESKY, that can take
// a BlockSparseMatrix as input.
return DENSE_QR;
}
if (linear_solver_type == ITERATIVE_SCHUR) {
return CGNR;
}
return linear_solver_type;
}
std::unique_ptr<LinearSolver> LinearSolver::Create(
const LinearSolver::Options& options) {
CHECK(options.context != nullptr);
switch (options.type) {
case CGNR: {
#ifndef CERES_NO_CUDA
if (options.sparse_linear_algebra_library_type == CUDA_SPARSE) {
std::string error;
return CudaCgnrSolver::Create(options, &error);
}
#endif
return std::make_unique<CgnrSolver>(options);
} break;
case SPARSE_NORMAL_CHOLESKY:
#if defined(CERES_NO_SPARSE)
return nullptr;
#else
if (options.dynamic_sparsity) {
return std::make_unique<DynamicSparseNormalCholeskySolver>(options);
}
return std::make_unique<SparseNormalCholeskySolver>(options);
#endif
case SPARSE_SCHUR:
#if defined(CERES_NO_SPARSE)
return nullptr;
#else
return std::make_unique<SparseSchurComplementSolver>(options);
#endif
case DENSE_SCHUR:
return std::make_unique<DenseSchurComplementSolver>(options);
case ITERATIVE_SCHUR:
if (options.use_explicit_schur_complement) {
return std::make_unique<SparseSchurComplementSolver>(options);
} else {
return std::make_unique<IterativeSchurComplementSolver>(options);
}
case DENSE_QR:
return std::make_unique<DenseQRSolver>(options);
case DENSE_NORMAL_CHOLESKY:
return std::make_unique<DenseNormalCholeskySolver>(options);
default:
LOG(FATAL) << "Unknown linear solver type :" << options.type;
return nullptr; // MSVC doesn't understand that LOG(FATAL) never returns.
}
}
} // namespace ceres::internal