)]}' { "commit": "42352e2e2a046673cc21e6c884817f5040899e6a", "tree": "52619fc47d1d278f9712beb522a50ea1db9a73b3", "parents": [ "f802a09ff12f79a3daaded70eaea2b033c0ce01f" ], "author": { "name": "Joydeep Biswas", "email": "joydeepb@cs.utexas.edu", "time": "Thu Aug 18 16:45:50 2022 -0500" }, "committer": { "name": "Joydeep Biswas", "email": "joydeepb@cs.utexas.edu", "time": "Fri Aug 26 08:57:07 2022 -0500" }, "message": "Added CUDA Jacobi Preconditioner\n\n* Added CudaCpuPreconditionerWrapper, a templated wrapper to leverage\n existing CPU preconditioners that provide CompressedRowSparse\n representation, for use with CudaCgnrSolver.\n* Added CudaJacobiPreconditioner as a specific instantiation of\n CudaCpuPreconditionerWrapper over BlockCRSJacobiPreconditioner.\n\nThe resulting Jacobi-preconditioned CudaCgnrSolver exhibits moderate\nincrease in speed with a healthy increase in numerical stability.\nExamples of running CudaCgnrSolver with Jacobi vs. Identity\npreconditioner for bundle adjustment on problem-13682-4456117-pre.txt\non a desktop computer with an Intel(R) Core(TM) i9-9940X CPU @ 3.30GHz\nand an Nvidia Quadro RTX 6000 GPU.\n\nA known issue is that the Jacobi preconditioner update is slow - this\ncan be sped up in the future by multithreading on the CPU, or by\ncomputing the preconditioner on the GPU.\n\n\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\nCUDA CGNR + IDENTITY Preconditioner\n\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\niter cost cost_change |gradient| |step| tr_ratio tr_radius ls_iter iter_time total_time\n 0 1.126372e+09 0.00e+00 3.02e+15 0.00e+00 0.00e+00 1.00e+04 0 9.17e+00 2.70e+01\n 1 3.796903e+07 1.09e+09 1.08e+15 6.48e+03 9.95e-01 3.00e+04 49 1.55e+01 4.24e+01\n 2 3.529024e+21 -3.53e+21 1.08e+15 1.93e+04 -2.25e+14 1.50e+04 500 3.79e+01 8.03e+01\n 3 7.952482e+09 -7.91e+09 1.08e+15 1.63e+04 -5.16e+02 3.75e+03 386 2.96e+01 1.10e+02\n 4 8.842366e+08 -8.46e+08 1.08e+15 1.18e+04 -5.87e+01 4.69e+02 312 2.47e+01 1.35e+02\n 5 2.935800e+07 8.61e+06 1.24e+13 3.93e+03 8.19e-01 6.32e+02 129 1.96e+01 1.54e+02\n 6 2.529342e+07 4.06e+06 3.22e+12 3.39e+03 9.93e-01 1.90e+03 148 2.16e+01 1.76e+02\n 7 2.343028e+07 1.86e+06 1.27e+13 6.00e+03 9.37e-01 5.68e+03 387 3.79e+01 2.14e+02\n 8 2.332405e+07 1.06e+05 7.92e+12 6.47e+01 8.26e-01 7.87e+03 46 1.48e+01 2.28e+02\n 9 3.591512e+14 -3.59e+14 7.92e+12 5.19e+03 -4.09e+08 3.94e+03 500 3.83e+01 2.67e+02\n 10 2.956078e+07 -6.24e+06 7.92e+12 3.94e+03 -8.34e+00 9.84e+02 447 3.40e+01 3.01e+02\n\nSolver Summary (v 2.2.0-eigen-(3.3.7)-lapack-suitesparse-(5.7.1)-metis-(5.1.0)-eigensparse-no_openmp-cuda-(11070))\n\n Original Reduced\nParameter blocks 4469799 4469799\nParameters 13491489 13491489\nResidual blocks 28987644 28987644\nResiduals 57975288 57975288\n\nMinimizer TRUST_REGION\nTrust region strategy LEVENBERG_MARQUARDT\nSparse linear algebra library CUDA_SPARSE\n\n Given Used\nLinear solver CGNR CGNR\nPreconditioner IDENTITY IDENTITY\nThreads 24 24\nLinear solver ordering 4456117,13682 4456117,13682\n\nCost:\nInitial 1.126372e+09\nFinal 2.332405e+07\nChange 1.103048e+09\n\nMinimizer iterations 11\nSuccessful steps 6\nUnsuccessful steps 5\n\nTime (in seconds):\nPreprocessor 17.798455\n\n Residual only evaluation 8.656350 (10)\n Jacobian \u0026 residual evaluation 35.794169 (6)\n Linear solver 208.586109 (10)\nMinimizer 283.593882\n\nPostprocessor 1.985112\nTotal 303.377449\n\n\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\nCUDA CGNR + JACOBI Preconditioner\n\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\niter cost cost_change |gradient| |step| tr_ratio tr_radius ls_iter iter_time total_time\n 0 1.126372e+09 0.00e+00 3.02e+15 0.00e+00 0.00e+00 1.00e+04 0 9.18e+00 2.68e+01\n 1 1.491225e+14 -1.49e+14 3.02e+15 8.73e+05 -1.36e+05 5.00e+03 6 1.07e+01 3.74e+01\n 2 7.461790e+12 -7.46e+12 3.02e+15 4.78e+05 -6.78e+03 1.25e+03 6 8.88e+00 4.63e+01\n 3 4.357438e+07 1.08e+09 5.41e+13 1.57e+05 9.87e-01 3.75e+03 6 1.62e+01 6.25e+01\n 4 2.928453e+07 1.43e+07 6.85e+13 3.81e+05 6.97e-01 3.99e+03 23 1.82e+01 8.08e+01\n 5 5.977433e+14 -5.98e+14 6.85e+13 4.97e+05 -8.70e+07 2.00e+03 9 9.93e+00 9.07e+01\n 6 6.423965e+10 -6.42e+10 6.85e+13 2.45e+05 -9.61e+03 4.99e+02 7 8.95e+00 9.97e+01\n 7 2.337471e+07 5.91e+06 1.49e+12 8.24e+04 9.31e-01 1.39e+03 4 1.60e+01 1.16e+02\n 8 2.257524e+07 7.99e+05 8.71e+12 1.62e+05 8.99e-01 2.81e+03 54 2.03e+01 1.36e+02\n 9 2.285090e+07 -2.76e+05 8.71e+12 3.56e+05 -6.13e-01 1.41e+03 104 1.65e+01 1.52e+02\n 10 5.909983e+12 -5.91e+12 8.71e+12 1.78e+05 -1.83e+07 3.52e+02 74 1.36e+01 1.66e+02\n\nSolver Summary (v 2.2.0-eigen-(3.3.7)-lapack-suitesparse-(5.7.1)-metis-(5.1.0)-eigensparse-no_openmp-cuda-(11070))\n\n Original Reduced\nParameter blocks 4469799 4469799\nParameters 13491489 13491489\nResidual blocks 28987644 28987644\nResiduals 57975288 57975288\n\nMinimizer TRUST_REGION\nTrust region strategy LEVENBERG_MARQUARDT\nSparse linear algebra library CUDA_SPARSE\n\n Given Used\nLinear solver CGNR CGNR\nPreconditioner JACOBI JACOBI\nThreads 24 24\nLinear solver ordering 4456117,13682 4456117,13682\n\nCost:\nInitial 1.126372e+09\nFinal 2.257524e+07\nChange 1.103796e+09\n\nMinimizer iterations 11\nSuccessful steps 5\nUnsuccessful steps 6\n\nTime (in seconds):\nPreprocessor 17.575990\n\n Residual only evaluation 8.461414 (10)\n Jacobian \u0026 residual evaluation 30.055383 (5)\n Linear solver 82.550675 (10)\nMinimizer 149.137856\n\nPostprocessor 1.994014\nTotal 168.707860\n\nChange-Id: I458d2445bf062e54de44fc91517ed11a300c7182\n", "tree_diff": [ { "type": "modify", "old_id": "5374f5103ae099ffb0da2a03e3c5431144d2a5ee", "old_mode": 33188, "old_path": "internal/ceres/cgnr_solver.cc", "new_id": "015e170951880197daded8945111418507fe92c3", "new_mode": 33188, "new_path": "internal/ceres/cgnr_solver.cc" }, { "type": "modify", "old_id": "9078d011e3c8b797798c54a6eb1db46c9c564797", "old_mode": 33188, "old_path": "internal/ceres/cgnr_solver.h", "new_id": "5111d808cba8b0eff91053e2f9b53436b8cb37f7", "new_mode": 33188, "new_path": "internal/ceres/cgnr_solver.h" }, { "type": "modify", "old_id": "0656d34c7d7a49a992746434661a4e7c4ead375d", "old_mode": 33188, "old_path": "internal/ceres/solver.cc", "new_id": "c896e329075373161e15dd7698fe342bdead3a80", "new_mode": 33188, "new_path": "internal/ceres/solver.cc" }, { "type": "modify", "old_id": "b796873952011decb3f63301fc5a9de1099823ec", "old_mode": 33188, "old_path": "internal/ceres/solver_test.cc", "new_id": "929e293f01e430033bbaf2aa713ce0c2d0c0607d", "new_mode": 33188, "new_path": "internal/ceres/solver_test.cc" } ] }