xref: /petsc/src/snes/tutorials/output/ex12_tri_parmetis_hpddm.out (revision 6a98f8dc3f2c9149905a87dc2e9d0fedaf64e09a)
1  0 SNES Function norm 33.3967
2    0 KSP Residual norm 60.6931
3    1 KSP Residual norm 17.0775
4    2 KSP Residual norm 3.02313
5    3 KSP Residual norm 0.829536
6    4 KSP Residual norm 0.194913
7    5 KSP Residual norm 0.0639834
8    6 KSP Residual norm 0.0197329
9    7 KSP Residual norm 0.00771292
10    8 KSP Residual norm 0.00275753
11    9 KSP Residual norm 0.000600238
12   10 KSP Residual norm 0.000149587
13   11 KSP Residual norm 3.29519e-05
14   12 KSP Residual norm 8.72355e-06
15   13 KSP Residual norm 1.91575e-06
16   14 KSP Residual norm 4.64281e-07
17   15 KSP Residual norm 8.45902e-08
18   16 KSP Residual norm 1.55073e-08
19   17 KSP Residual norm 3.3689e-09
20  Linear solve converged due to CONVERGED_RTOL iterations 17
21  1 SNES Function norm 3.34156e-09
22L_2 Error: 7.89093
23Nonlinear solve converged due to CONVERGED_FNORM_RELATIVE iterations 1
24SNES Object: 4 MPI processes
25  type: newtonls
26  maximum iterations=50, maximum function evaluations=10000
27  tolerances: relative=1e-08, absolute=1e-50, solution=1e-08
28  total number of linear solver iterations=17
29  total number of function evaluations=2
30  norm schedule ALWAYS
31  SNESLineSearch Object: 4 MPI processes
32    type: bt
33      interpolation: cubic
34      alpha=1.000000e-04
35    maxstep=1.000000e+08, minlambda=1.000000e-12
36    tolerances: relative=1.000000e-08, absolute=1.000000e-15, lambda=1.000000e-08
37    maximum iterations=40
38  KSP Object: 4 MPI processes
39    type: gmres
40      restart=100, using Classical (unmodified) Gram-Schmidt Orthogonalization with no iterative refinement
41      happy breakdown tolerance 1e-30
42    maximum iterations=10000, initial guess is zero
43    tolerances:  relative=1e-10, absolute=1e-50, divergence=10000.
44    left preconditioning
45    using PRECONDITIONED norm type for convergence test
46  PC Object: 4 MPI processes
47    type: hpddm
48    levels: 2
49    Neumann matrix attached? TRUE
50    coarse correction: deflated
51    on process #0, value (+ threshold if available) for selecting deflation vectors: 20
52    grid and operator complexities: 1.01463 1.10782
53    KSP Object: (pc_hpddm_levels_1_) 4 MPI processes
54      type: preonly
55      maximum iterations=10000, initial guess is zero
56      tolerances:  relative=1e-05, absolute=1e-50, divergence=10000.
57      left preconditioning
58      using NONE norm type for convergence test
59    PC Object: (pc_hpddm_levels_1_) 4 MPI processes
60      type: bjacobi
61        number of blocks = 4
62        Local solve is same for all blocks, in the following KSP and PC objects:
63      KSP Object: (pc_hpddm_levels_1_sub_) 1 MPI processes
64        type: preonly
65        maximum iterations=10000, initial guess is zero
66        tolerances:  relative=1e-05, absolute=1e-50, divergence=10000.
67        left preconditioning
68        using NONE norm type for convergence test
69      PC Object: (pc_hpddm_levels_1_sub_) 1 MPI processes
70        type: icc
71          out-of-place factorization
72          3 levels of fill
73          tolerance for zero pivot 2.22045e-14
74          using Manteuffel shift [POSITIVE_DEFINITE]
75          matrix ordering: natural
76          factor fill ratio given 1., needed 2.69786
77            Factored matrix follows:
78              Mat Object: 1 MPI processes
79                type: seqsbaij
80                rows=1301, cols=1301
81                package used to perform factorization: petsc
82                total: nonzeros=13635, allocated nonzeros=13635
83                total number of mallocs used during MatSetValues calls=0
84                    block size is 1
85        linear system matrix = precond matrix:
86        Mat Object: 1 MPI processes
87          type: seqaij
88          rows=1301, cols=1301
89          total: nonzeros=8807, allocated nonzeros=8807
90          total number of mallocs used during MatSetValues calls=0
91            not using I-node routines
92      linear system matrix = precond matrix:
93      Mat Object: 4 MPI processes
94        type: mpiaij
95        rows=5468, cols=5468
96        total: nonzeros=37098, allocated nonzeros=37098
97        total number of mallocs used during MatSetValues calls=0
98          not using I-node (on process 0) routines
99      KSP Object: (pc_hpddm_coarse_) 2 MPI processes
100        type: preonly
101        maximum iterations=10000, initial guess is zero
102        tolerances:  relative=1e-05, absolute=1e-50, divergence=10000.
103        left preconditioning
104        using NONE norm type for convergence test
105      PC Object: (pc_hpddm_coarse_) 2 MPI processes
106        type: redundant
107          First (color=0) of 2 PCs follows
108        linear system matrix = precond matrix:
109        Mat Object: (pc_hpddm_coarse_) 2 MPI processes
110          type: mpisbaij
111          rows=80, cols=80, bs=20
112          total: nonzeros=4000, allocated nonzeros=4000
113          total number of mallocs used during MatSetValues calls=0
114              block size is 20
115                KSP Object:       (pc_hpddm_coarse_redundant_)       1 MPI processes
116                  type: preonly
117                  maximum iterations=10000, initial guess is zero
118                  tolerances:  relative=1e-05, absolute=1e-50, divergence=10000.
119                  left preconditioning
120                  using NONE norm type for convergence test
121                PC Object:       (pc_hpddm_coarse_redundant_)       1 MPI processes
122                  type: cholesky
123                    out-of-place factorization
124                    tolerance for zero pivot 2.22045e-14
125                    matrix ordering: natural
126                    factor fill ratio given 5., needed 1.1
127                      Factored matrix follows:
128                        Mat Object:       1 MPI processes
129                          type: seqsbaij
130                          rows=80, cols=80, bs=20
131                          package used to perform factorization: petsc
132                          total: nonzeros=4400, allocated nonzeros=4400
133                          total number of mallocs used during MatSetValues calls=0
134                              block size is 20
135                  linear system matrix = precond matrix:
136                  Mat Object:       1 MPI processes
137                    type: seqsbaij
138                    rows=80, cols=80, bs=20
139                    total: nonzeros=4000, allocated nonzeros=4000
140                    total number of mallocs used during MatSetValues calls=0
141                        block size is 20
142    linear system matrix = precond matrix:
143    Mat Object: 4 MPI processes
144      type: mpiaij
145      rows=5468, cols=5468
146      total: nonzeros=37098, allocated nonzeros=37098
147      total number of mallocs used during MatSetValues calls=0
148        not using I-node (on process 0) routines
149