======================== JOB MAP ======================== Data for node: Name: tesla2 Num procs: 4 Process OMPI jobid: [48135,1] Process rank: 0 Process OMPI jobid: [48135,1] Process rank: 1 Process OMPI jobid: [48135,1] Process rank: 2 Process OMPI jobid: [48135,1] Process rank: 3 ============================================================= ================================================================================ HPLinpack 2.0 -- High-Performance Linpack benchmark -- September 10, 2008 Written by A. Petitet and R. Clint Whaley, Innovative Computing Laboratory, UTK Modified by Piotr Luszczek, Innovative Computing Laboratory, UTK Modified by Julien Langou, University of Colorado Denver ================================================================================ An explanation of the input/output parameters follows: T/V : Wall time / encoded variant. N : The order of the coefficient matrix A. NB : The partitioning blocking factor. P : The number of process rows. Q : The number of process columns. Time : Time in seconds to solve the linear system. Gflops : Rate of execution for solving the linear system. The following parameter values will be used: N : 6000 NB : 960 PMAP : Row-major process mapping P : 2 Q : 2 PFACT : Left NBMIN : 2 NDIV : 4 RFACT : Left BCAST : 2ringM DEPTH : 0 SWAP : Binary-exchange L1 : no-transposed form U : no-transposed form EQUIL : no ALIGN : 8 double precision words -------------------------------------------------------------------------------- - The matrix A is randomly generated for each test. - The following scaled residual check will be computed: ||Ax-b||_oo / ( eps * ( || x ||_oo * || A ||_oo + || b ||_oo ) * N ) - The relative machine precision (eps) is taken to be 1.110223e-16 - Computational tests pass if scaled residuals are less than 16.0 Assigning device 0 to process on node tesla2.localdomain rank 0 Assigning device 2 to process on node tesla2.localdomain rank 2 Assigning device 1 to process on node tesla2.localdomain rank 1 Assigning device 3 to process on node tesla2.localdomain rank 3 DTRSM split from environment variable 0.650000 DTRSM split from environment variable 0.650000 DGEMM split from environment variable 0.800000 DGEMM split from environment variable 0.800000 DTRSM split from environment variable 0.650000 DTRSM split from environment variable 0.650000 DGEMM split from environment variable 0.800000 DGEMM split from environment variable 0.800000 ================================================================================ T/V N NB P Q Time Gflops -------------------------------------------------------------------------------- WR03L4L2 6000 960 2 2 2.98 4.828e+01 -------------------------------------------------------------------------------- ||Ax-b||_oo/(eps*(||A||_oo*||x||_oo+||b||_oo)*N)= 0.0050869 ...... PASSED ================================================================================ Finished 1 tests with the following results: 1 tests completed and passed residual checks, 0 tests completed and failed residual checks, 0 tests skipped because of illegal input values. -------------------------------------------------------------------------------- End of Tests. ================================================================================ ======================== JOB MAP ======================== Data for node: Name: tesla2 Num procs: 4 Process OMPI jobid: [48377,1] Process rank: 0 Process OMPI jobid: [48377,1] Process rank: 1 Process OMPI jobid: [48377,1] Process rank: 2 Process OMPI jobid: [48377,1] Process rank: 3 ============================================================= ================================================================================ HPLinpack 2.0 -- High-Performance Linpack benchmark -- September 10, 2008 Written by A. Petitet and R. Clint Whaley, Innovative Computing Laboratory, UTK Modified by Piotr Luszczek, Innovative Computing Laboratory, UTK Modified by Julien Langou, University of Colorado Denver ================================================================================ An explanation of the input/output parameters follows: T/V : Wall time / encoded variant. N : The order of the coefficient matrix A. NB : The partitioning blocking factor. P : The number of process rows. Q : The number of process columns. Time : Time in seconds to solve the linear system. Gflops : Rate of execution for solving the linear system. The following parameter values will be used: N : 6000 NB : 960 PMAP : Row-major process mapping P : 2 Q : 2 PFACT : Left NBMIN : 2 NDIV : 4 RFACT : Left BCAST : 2ringM DEPTH : 0 SWAP : Binary-exchange L1 : no-transposed form U : no-transposed form EQUIL : no ALIGN : 8 double precision words -------------------------------------------------------------------------------- - The matrix A is randomly generated for each test. - The following scaled residual check will be computed: ||Ax-b||_oo / ( eps * ( || x ||_oo * || A ||_oo + || b ||_oo ) * N ) - The relative machine precision (eps) is taken to be 1.110223e-16 - Computational tests pass if scaled residuals are less than 16.0 Assigning device 0 to process on node tesla2.localdomain rank 0 Assigning device 1 to process on node tesla2.localdomain rank 1 Assigning device 2 to process on node tesla2.localdomain rank 2 Assigning device 3 to process on node tesla2.localdomain rank 3 DTRSM split from environment variable 0.650000 DTRSM split from environment variable 0.650000 DGEMM split from environment variable 0.500000 DGEMM split from environment variable 0.500000 DTRSM split from environment variable 0.650000 DTRSM split from environment variable 0.650000 DGEMM split from environment variable 0.500000 DGEMM split from environment variable 0.500000 ================================================================================ T/V N NB P Q Time Gflops -------------------------------------------------------------------------------- WR03L4L2 6000 960 2 2 3.58 4.028e+01 -------------------------------------------------------------------------------- ||Ax-b||_oo/(eps*(||A||_oo*||x||_oo+||b||_oo)*N)= 0.0060513 ...... PASSED ================================================================================ Finished 1 tests with the following results: 1 tests completed and passed residual checks, 0 tests completed and failed residual checks, 0 tests skipped because of illegal input values. -------------------------------------------------------------------------------- End of Tests. ================================================================================ ======================== JOB MAP ======================== Data for node: Name: tesla2 Num procs: 4 Process OMPI jobid: [48371,1] Process rank: 0 Process OMPI jobid: [48371,1] Process rank: 1 Process OMPI jobid: [48371,1] Process rank: 2 Process OMPI jobid: [48371,1] Process rank: 3 ============================================================= ================================================================================ HPLinpack 2.0 -- High-Performance Linpack benchmark -- September 10, 2008 Written by A. Petitet and R. Clint Whaley, Innovative Computing Laboratory, UTK Modified by Piotr Luszczek, Innovative Computing Laboratory, UTK Modified by Julien Langou, University of Colorado Denver ================================================================================ An explanation of the input/output parameters follows: T/V : Wall time / encoded variant. N : The order of the coefficient matrix A. NB : The partitioning blocking factor. P : The number of process rows. Q : The number of process columns. Time : Time in seconds to solve the linear system. Gflops : Rate of execution for solving the linear system. The following parameter values will be used: N : 6000 NB : 960 PMAP : Row-major process mapping P : 2 Q : 2 PFACT : Left NBMIN : 2 NDIV : 4 RFACT : Left BCAST : 2ringM DEPTH : 0 SWAP : Binary-exchange L1 : no-transposed form U : no-transposed form EQUIL : no ALIGN : 8 double precision words -------------------------------------------------------------------------------- - The matrix A is randomly generated for each test. - The following scaled residual check will be computed: ||Ax-b||_oo / ( eps * ( || x ||_oo * || A ||_oo + || b ||_oo ) * N ) - The relative machine precision (eps) is taken to be 1.110223e-16 - Computational tests pass if scaled residuals are less than 16.0 Assigning device 0 to process on node tesla2.localdomain rank 0 Assigning device 2 to process on node tesla2.localdomain rank 2 Assigning device 3 to process on node tesla2.localdomain rank 3 Assigning device 1 to process on node tesla2.localdomain rank 1 DTRSM split from environment variable 0.650000 DTRSM split from environment variable 0.650000 DGEMM split from environment variable 0.750000 DGEMM split from environment variable 0.750000 DTRSM split from environment variable 0.650000 DGEMM split from environment variable 0.750000 DTRSM split from environment variable 0.650000 DGEMM split from environment variable 0.750000 ================================================================================ T/V N NB P Q Time Gflops -------------------------------------------------------------------------------- WR03L4L2 6000 960 2 2 3.10 4.652e+01 -------------------------------------------------------------------------------- ||Ax-b||_oo/(eps*(||A||_oo*||x||_oo+||b||_oo)*N)= 0.0054978 ...... PASSED ================================================================================ Finished 1 tests with the following results: 1 tests completed and passed residual checks, 0 tests completed and failed residual checks, 0 tests skipped because of illegal input values. -------------------------------------------------------------------------------- End of Tests. ================================================================================ ======================== JOB MAP ======================== Data for node: Name: tesla2 Num procs: 4 Process OMPI jobid: [48363,1] Process rank: 0 Process OMPI jobid: [48363,1] Process rank: 1 Process OMPI jobid: [48363,1] Process rank: 2 Process OMPI jobid: [48363,1] Process rank: 3 ============================================================= ================================================================================ HPLinpack 2.0 -- High-Performance Linpack benchmark -- September 10, 2008 Written by A. Petitet and R. Clint Whaley, Innovative Computing Laboratory, UTK Modified by Piotr Luszczek, Innovative Computing Laboratory, UTK Modified by Julien Langou, University of Colorado Denver ================================================================================ An explanation of the input/output parameters follows: T/V : Wall time / encoded variant. N : The order of the coefficient matrix A. NB : The partitioning blocking factor. P : The number of process rows. Q : The number of process columns. Time : Time in seconds to solve the linear system. Gflops : Rate of execution for solving the linear system. The following parameter values will be used: N : 6000 NB : 960 PMAP : Row-major process mapping P : 2 Q : 2 PFACT : Left NBMIN : 2 NDIV : 4 RFACT : Left BCAST : 2ringM DEPTH : 0 SWAP : Binary-exchange L1 : no-transposed form U : no-transposed form EQUIL : no ALIGN : 8 double precision words -------------------------------------------------------------------------------- - The matrix A is randomly generated for each test. - The following scaled residual check will be computed: ||Ax-b||_oo / ( eps * ( || x ||_oo * || A ||_oo + || b ||_oo ) * N ) - The relative machine precision (eps) is taken to be 1.110223e-16 - Computational tests pass if scaled residuals are less than 16.0 Assigning device 0 to process on node tesla2.localdomain rank 0 Assigning device 1 to process on node tesla2.localdomain rank 1 Assigning device 2 to process on node tesla2.localdomain rank 2 Assigning device 3 to process on node tesla2.localdomain rank 3 DTRSM split from environment variable 0.650000 DTRSM split from environment variable 0.650000 DGEMM split from environment variable 0.790000 DGEMM split from environment variable 0.790000 DTRSM split from environment variable 0.650000 DGEMM split from environment variable 0.790000 DTRSM split from environment variable 0.650000 DGEMM split from environment variable 0.790000 ================================================================================ T/V N NB P Q Time Gflops -------------------------------------------------------------------------------- WR03L4L2 6000 960 2 2 3.02 4.777e+01 -------------------------------------------------------------------------------- ||Ax-b||_oo/(eps*(||A||_oo*||x||_oo+||b||_oo)*N)= 0.0047623 ...... PASSED ================================================================================ Finished 1 tests with the following results: 1 tests completed and passed residual checks, 0 tests completed and failed residual checks, 0 tests skipped because of illegal input values. -------------------------------------------------------------------------------- End of Tests. ================================================================================ ======================== JOB MAP ======================== Data for node: Name: tesla2 Num procs: 4 Process OMPI jobid: [48355,1] Process rank: 0 Process OMPI jobid: [48355,1] Process rank: 1 Process OMPI jobid: [48355,1] Process rank: 2 Process OMPI jobid: [48355,1] Process rank: 3 ============================================================= ================================================================================ HPLinpack 2.0 -- High-Performance Linpack benchmark -- September 10, 2008 Written by A. Petitet and R. Clint Whaley, Innovative Computing Laboratory, UTK Modified by Piotr Luszczek, Innovative Computing Laboratory, UTK Modified by Julien Langou, University of Colorado Denver ================================================================================ An explanation of the input/output parameters follows: T/V : Wall time / encoded variant. N : The order of the coefficient matrix A. NB : The partitioning blocking factor. P : The number of process rows. Q : The number of process columns. Time : Time in seconds to solve the linear system. Gflops : Rate of execution for solving the linear system. The following parameter values will be used: N : 6000 NB : 960 PMAP : Row-major process mapping P : 2 Q : 2 PFACT : Left NBMIN : 2 NDIV : 4 RFACT : Left BCAST : 2ringM DEPTH : 0 SWAP : Binary-exchange L1 : no-transposed form U : no-transposed form EQUIL : no ALIGN : 8 double precision words -------------------------------------------------------------------------------- - The matrix A is randomly generated for each test. - The following scaled residual check will be computed: ||Ax-b||_oo / ( eps * ( || x ||_oo * || A ||_oo + || b ||_oo ) * N ) - The relative machine precision (eps) is taken to be 1.110223e-16 - Computational tests pass if scaled residuals are less than 16.0 Assigning device 0 to process on node tesla2.localdomain rank 0 Assigning device 3 to process on node tesla2.localdomain rank 3 Assigning device 1 to process on node tesla2.localdomain rank 1 Assigning device 2 to process on node tesla2.localdomain rank 2 DTRSM split from environment variable 0.550000 DTRSM split from environment variable 0.550000 DGEMM split from environment variable 0.790000 DGEMM split from environment variable 0.790000 DTRSM split from environment variable 0.550000 DTRSM split from environment variable 0.550000 DGEMM split from environment variable 0.790000 DGEMM split from environment variable 0.790000 ================================================================================ T/V N NB P Q Time Gflops -------------------------------------------------------------------------------- WR03L4L2 6000 960 2 2 3.08 4.683e+01 -------------------------------------------------------------------------------- ||Ax-b||_oo/(eps*(||A||_oo*||x||_oo+||b||_oo)*N)= 0.0051655 ...... PASSED ================================================================================ Finished 1 tests with the following results: 1 tests completed and passed residual checks, 0 tests completed and failed residual checks, 0 tests skipped because of illegal input values. -------------------------------------------------------------------------------- End of Tests. ================================================================================ ======================== JOB MAP ======================== Data for node: Name: tesla2 Num procs: 4 Process OMPI jobid: [48341,1] Process rank: 0 Process OMPI jobid: [48341,1] Process rank: 1 Process OMPI jobid: [48341,1] Process rank: 2 Process OMPI jobid: [48341,1] Process rank: 3 ============================================================= ================================================================================ HPLinpack 2.0 -- High-Performance Linpack benchmark -- September 10, 2008 Written by A. Petitet and R. Clint Whaley, Innovative Computing Laboratory, UTK Modified by Piotr Luszczek, Innovative Computing Laboratory, UTK Modified by Julien Langou, University of Colorado Denver ================================================================================ An explanation of the input/output parameters follows: T/V : Wall time / encoded variant. N : The order of the coefficient matrix A. NB : The partitioning blocking factor. P : The number of process rows. Q : The number of process columns. Time : Time in seconds to solve the linear system. Gflops : Rate of execution for solving the linear system. The following parameter values will be used: N : 6000 NB : 960 PMAP : Row-major process mapping P : 2 Q : 2 PFACT : Left NBMIN : 2 NDIV : 4 RFACT : Left BCAST : 2ringM DEPTH : 0 SWAP : Binary-exchange L1 : no-transposed form U : no-transposed form EQUIL : no ALIGN : 8 double precision words -------------------------------------------------------------------------------- - The matrix A is randomly generated for each test. - The following scaled residual check will be computed: ||Ax-b||_oo / ( eps * ( || x ||_oo * || A ||_oo + || b ||_oo ) * N ) - The relative machine precision (eps) is taken to be 1.110223e-16 - Computational tests pass if scaled residuals are less than 16.0 Assigning device 0 to process on node tesla2.localdomain rank 0 Assigning device 2 to process on node tesla2.localdomain rank 2 Assigning device 1 to process on node tesla2.localdomain rank 1 Assigning device 3 to process on node tesla2.localdomain rank 3 DTRSM split from environment variable 0.750000 DTRSM split from environment variable 0.750000 DGEMM split from environment variable 0.790000 DGEMM split from environment variable 0.790000 DTRSM split from environment variable 0.750000 DGEMM split from environment variable 0.790000 DTRSM split from environment variable 0.750000 DGEMM split from environment variable 0.790000 ================================================================================ T/V N NB P Q Time Gflops -------------------------------------------------------------------------------- WR03L4L2 6000 960 2 2 2.94 4.895e+01 -------------------------------------------------------------------------------- ||Ax-b||_oo/(eps*(||A||_oo*||x||_oo+||b||_oo)*N)= 0.0049823 ...... PASSED ================================================================================ Finished 1 tests with the following results: 1 tests completed and passed residual checks, 0 tests completed and failed residual checks, 0 tests skipped because of illegal input values. -------------------------------------------------------------------------------- End of Tests. ================================================================================ ======================== JOB MAP ======================== Data for node: Name: tesla2 Num procs: 4 Process OMPI jobid: [48333,1] Process rank: 0 Process OMPI jobid: [48333,1] Process rank: 1 Process OMPI jobid: [48333,1] Process rank: 2 Process OMPI jobid: [48333,1] Process rank: 3 ============================================================= ================================================================================ HPLinpack 2.0 -- High-Performance Linpack benchmark -- September 10, 2008 Written by A. Petitet and R. Clint Whaley, Innovative Computing Laboratory, UTK Modified by Piotr Luszczek, Innovative Computing Laboratory, UTK Modified by Julien Langou, University of Colorado Denver ================================================================================ An explanation of the input/output parameters follows: T/V : Wall time / encoded variant. N : The order of the coefficient matrix A. NB : The partitioning blocking factor. P : The number of process rows. Q : The number of process columns. Time : Time in seconds to solve the linear system. Gflops : Rate of execution for solving the linear system. The following parameter values will be used: N : 6000 NB : 960 PMAP : Row-major process mapping P : 2 Q : 2 PFACT : Left NBMIN : 2 NDIV : 4 RFACT : Left BCAST : 2ringM DEPTH : 0 SWAP : Binary-exchange L1 : no-transposed form U : no-transposed form EQUIL : no ALIGN : 8 double precision words -------------------------------------------------------------------------------- - The matrix A is randomly generated for each test. - The following scaled residual check will be computed: ||Ax-b||_oo / ( eps * ( || x ||_oo * || A ||_oo + || b ||_oo ) * N ) - The relative machine precision (eps) is taken to be 1.110223e-16 - Computational tests pass if scaled residuals are less than 16.0 Assigning device 0 to process on node tesla2.localdomain rank 0 Assigning device 1 to process on node tesla2.localdomain rank 1 Assigning device 2 to process on node tesla2.localdomain rank 2 Assigning device 3 to process on node tesla2.localdomain rank 3 DTRSM split from environment variable 0.750000 DTRSM split from environment variable 0.750000 DGEMM split from environment variable 0.800000 DGEMM split from environment variable 0.800000 DTRSM split from environment variable 0.750000 DGEMM split from environment variable 0.800000 DTRSM split from environment variable 0.750000 DGEMM split from environment variable 0.800000 ================================================================================ T/V N NB P Q Time Gflops -------------------------------------------------------------------------------- WR03L4L2 6000 960 2 2 2.93 4.912e+01 -------------------------------------------------------------------------------- ||Ax-b||_oo/(eps*(||A||_oo*||x||_oo+||b||_oo)*N)= 0.0054644 ...... PASSED ================================================================================ Finished 1 tests with the following results: 1 tests completed and passed residual checks, 0 tests completed and failed residual checks, 0 tests skipped because of illegal input values. -------------------------------------------------------------------------------- End of Tests. ================================================================================ ======================== JOB MAP ======================== Data for node: Name: tesla2 Num procs: 4 Process OMPI jobid: [48325,1] Process rank: 0 Process OMPI jobid: [48325,1] Process rank: 1 Process OMPI jobid: [48325,1] Process rank: 2 Process OMPI jobid: [48325,1] Process rank: 3 ============================================================= ================================================================================ HPLinpack 2.0 -- High-Performance Linpack benchmark -- September 10, 2008 Written by A. Petitet and R. Clint Whaley, Innovative Computing Laboratory, UTK Modified by Piotr Luszczek, Innovative Computing Laboratory, UTK Modified by Julien Langou, University of Colorado Denver ================================================================================ An explanation of the input/output parameters follows: T/V : Wall time / encoded variant. N : The order of the coefficient matrix A. NB : The partitioning blocking factor. P : The number of process rows. Q : The number of process columns. Time : Time in seconds to solve the linear system. Gflops : Rate of execution for solving the linear system. The following parameter values will be used: N : 6000 NB : 960 PMAP : Row-major process mapping P : 2 Q : 2 PFACT : Left NBMIN : 2 NDIV : 4 RFACT : Left BCAST : 2ringM DEPTH : 0 SWAP : Binary-exchange L1 : no-transposed form U : no-transposed form EQUIL : no ALIGN : 8 double precision words -------------------------------------------------------------------------------- - The matrix A is randomly generated for each test. - The following scaled residual check will be computed: ||Ax-b||_oo / ( eps * ( || x ||_oo * || A ||_oo + || b ||_oo ) * N ) - The relative machine precision (eps) is taken to be 1.110223e-16 - Computational tests pass if scaled residuals are less than 16.0 Assigning device 0 to process on node tesla2.localdomain rank 0 Assigning device 1 to process on node tesla2.localdomain rank 1 Assigning device 2 to process on node tesla2.localdomain rank 2 Assigning device 3 to process on node tesla2.localdomain rank 3 DTRSM split from environment variable 0.770000 DTRSM split from environment variable 0.770000 DGEMM split from environment variable 0.800000 DGEMM split from environment variable 0.800000 DTRSM split from environment variable 0.770000 DGEMM split from environment variable 0.800000 DTRSM split from environment variable 0.770000 DGEMM split from environment variable 0.800000 ================================================================================ T/V N NB P Q Time Gflops -------------------------------------------------------------------------------- WR03L4L2 6000 960 2 2 3.01 4.783e+01 -------------------------------------------------------------------------------- ||Ax-b||_oo/(eps*(||A||_oo*||x||_oo+||b||_oo)*N)= 0.0048209 ...... PASSED ================================================================================ Finished 1 tests with the following results: 1 tests completed and passed residual checks, 0 tests completed and failed residual checks, 0 tests skipped because of illegal input values. -------------------------------------------------------------------------------- End of Tests. ================================================================================ ======================== JOB MAP ======================== Data for node: Name: tesla2 Num procs: 4 Process OMPI jobid: [48313,1] Process rank: 0 Process OMPI jobid: [48313,1] Process rank: 1 Process OMPI jobid: [48313,1] Process rank: 2 Process OMPI jobid: [48313,1] Process rank: 3 ============================================================= ================================================================================ HPLinpack 2.0 -- High-Performance Linpack benchmark -- September 10, 2008 Written by A. Petitet and R. Clint Whaley, Innovative Computing Laboratory, UTK Modified by Piotr Luszczek, Innovative Computing Laboratory, UTK Modified by Julien Langou, University of Colorado Denver ================================================================================ An explanation of the input/output parameters follows: T/V : Wall time / encoded variant. N : The order of the coefficient matrix A. NB : The partitioning blocking factor. P : The number of process rows. Q : The number of process columns. Time : Time in seconds to solve the linear system. Gflops : Rate of execution for solving the linear system. The following parameter values will be used: N : 6000 NB : 960 PMAP : Row-major process mapping P : 2 Q : 2 PFACT : Left NBMIN : 2 NDIV : 4 RFACT : Left BCAST : 2ringM DEPTH : 0 SWAP : Binary-exchange L1 : no-transposed form U : no-transposed form EQUIL : no ALIGN : 8 double precision words -------------------------------------------------------------------------------- - The matrix A is randomly generated for each test. - The following scaled residual check will be computed: ||Ax-b||_oo / ( eps * ( || x ||_oo * || A ||_oo + || b ||_oo ) * N ) - The relative machine precision (eps) is taken to be 1.110223e-16 - Computational tests pass if scaled residuals are less than 16.0 Assigning device 0 to process on node tesla2.localdomain rank 0 Assigning device 3 to process on node tesla2.localdomain rank 3 Assigning device 1 to process on node tesla2.localdomain rank 1 Assigning device 2 to process on node tesla2.localdomain rank 2 DTRSM split from environment variable 0.750000 DTRSM split from environment variable 0.750000 DGEMM split from environment variable 0.900000 DGEMM split from environment variable 0.900000 DTRSM split from environment variable 0.750000 DGEMM split from environment variable 0.900000 DTRSM split from environment variable 0.750000 DGEMM split from environment variable 0.900000 ================================================================================ T/V N NB P Q Time Gflops -------------------------------------------------------------------------------- WR03L4L2 6000 960 2 2 2.90 4.969e+01 -------------------------------------------------------------------------------- ||Ax-b||_oo/(eps*(||A||_oo*||x||_oo+||b||_oo)*N)= 0.0061858 ...... PASSED ================================================================================ Finished 1 tests with the following results: 1 tests completed and passed residual checks, 0 tests completed and failed residual checks, 0 tests skipped because of illegal input values. -------------------------------------------------------------------------------- End of Tests. ================================================================================ ======================== JOB MAP ======================== Data for node: Name: tesla2 Num procs: 4 Process OMPI jobid: [48305,1] Process rank: 0 Process OMPI jobid: [48305,1] Process rank: 1 Process OMPI jobid: [48305,1] Process rank: 2 Process OMPI jobid: [48305,1] Process rank: 3 ============================================================= ================================================================================ HPLinpack 2.0 -- High-Performance Linpack benchmark -- September 10, 2008 Written by A. Petitet and R. Clint Whaley, Innovative Computing Laboratory, UTK Modified by Piotr Luszczek, Innovative Computing Laboratory, UTK Modified by Julien Langou, University of Colorado Denver ================================================================================ An explanation of the input/output parameters follows: T/V : Wall time / encoded variant. N : The order of the coefficient matrix A. NB : The partitioning blocking factor. P : The number of process rows. Q : The number of process columns. Time : Time in seconds to solve the linear system. Gflops : Rate of execution for solving the linear system. The following parameter values will be used: N : 6000 NB : 960 PMAP : Row-major process mapping P : 2 Q : 2 PFACT : Left NBMIN : 2 NDIV : 4 RFACT : Left BCAST : 2ringM DEPTH : 0 SWAP : Binary-exchange L1 : no-transposed form U : no-transposed form EQUIL : no ALIGN : 8 double precision words -------------------------------------------------------------------------------- - The matrix A is randomly generated for each test. - The following scaled residual check will be computed: ||Ax-b||_oo / ( eps * ( || x ||_oo * || A ||_oo + || b ||_oo ) * N ) - The relative machine precision (eps) is taken to be 1.110223e-16 - Computational tests pass if scaled residuals are less than 16.0 Assigning device 0 to process on node tesla2.localdomain rank 0 Assigning device 2 to process on node tesla2.localdomain rank 2 Assigning device 3 to process on node tesla2.localdomain rank 3 Assigning device 1 to process on node tesla2.localdomain rank 1 DTRSM split from environment variable 0.750000 DTRSM split from environment variable 0.750000 DGEMM split from environment variable 0.990000 DGEMM split from environment variable 0.990000 DTRSM split from environment variable 0.750000 DGEMM split from environment variable 0.990000 DTRSM split from environment variable 0.750000 DGEMM split from environment variable 0.990000 ================================================================================ T/V N NB P Q Time Gflops -------------------------------------------------------------------------------- WR03L4L2 6000 960 2 2 2.93 4.921e+01 -------------------------------------------------------------------------------- ||Ax-b||_oo/(eps*(||A||_oo*||x||_oo+||b||_oo)*N)= 0.0051073 ...... PASSED ================================================================================ Finished 1 tests with the following results: 1 tests completed and passed residual checks, 0 tests completed and failed residual checks, 0 tests skipped because of illegal input values. -------------------------------------------------------------------------------- End of Tests. ================================================================================ ======================== JOB MAP ======================== Data for node: Name: tesla2 Num procs: 4 Process OMPI jobid: [48297,1] Process rank: 0 Process OMPI jobid: [48297,1] Process rank: 1 Process OMPI jobid: [48297,1] Process rank: 2 Process OMPI jobid: [48297,1] Process rank: 3 ============================================================= ================================================================================ HPLinpack 2.0 -- High-Performance Linpack benchmark -- September 10, 2008 Written by A. Petitet and R. Clint Whaley, Innovative Computing Laboratory, UTK Modified by Piotr Luszczek, Innovative Computing Laboratory, UTK Modified by Julien Langou, University of Colorado Denver ================================================================================ An explanation of the input/output parameters follows: T/V : Wall time / encoded variant. N : The order of the coefficient matrix A. NB : The partitioning blocking factor. P : The number of process rows. Q : The number of process columns. Time : Time in seconds to solve the linear system. Gflops : Rate of execution for solving the linear system. The following parameter values will be used: N : 6000 NB : 960 PMAP : Row-major process mapping P : 2 Q : 2 PFACT : Left NBMIN : 2 NDIV : 4 RFACT : Left BCAST : 2ringM DEPTH : 0 SWAP : Binary-exchange L1 : no-transposed form U : no-transposed form EQUIL : no ALIGN : 8 double precision words -------------------------------------------------------------------------------- - The matrix A is randomly generated for each test. - The following scaled residual check will be computed: ||Ax-b||_oo / ( eps * ( || x ||_oo * || A ||_oo + || b ||_oo ) * N ) - The relative machine precision (eps) is taken to be 1.110223e-16 - Computational tests pass if scaled residuals are less than 16.0 Assigning device 0 to process on node tesla2.localdomain rank 0 Assigning device 2 to process on node tesla2.localdomain rank 2 Assigning device 3 to process on node tesla2.localdomain rank 3 Assigning device 1 to process on node tesla2.localdomain rank 1 DTRSM split from environment variable 0.650000 DTRSM split from environment variable 0.650000 DGEMM split from environment variable 0.900000 DGEMM split from environment variable 0.900000 DTRSM split from environment variable 0.650000 DGEMM split from environment variable 0.900000 DTRSM split from environment variable 0.650000 DGEMM split from environment variable 0.900000 ================================================================================ T/V N NB P Q Time Gflops -------------------------------------------------------------------------------- WR03L4L2 6000 960 2 2 2.92 4.940e+01 -------------------------------------------------------------------------------- ||Ax-b||_oo/(eps*(||A||_oo*||x||_oo+||b||_oo)*N)= 0.0059279 ...... PASSED ================================================================================ Finished 1 tests with the following results: 1 tests completed and passed residual checks, 0 tests completed and failed residual checks, 0 tests skipped because of illegal input values. -------------------------------------------------------------------------------- End of Tests. ================================================================================ ======================== JOB MAP ======================== Data for node: Name: tesla2 Num procs: 4 Process OMPI jobid: [48291,1] Process rank: 0 Process OMPI jobid: [48291,1] Process rank: 1 Process OMPI jobid: [48291,1] Process rank: 2 Process OMPI jobid: [48291,1] Process rank: 3 ============================================================= ================================================================================ HPLinpack 2.0 -- High-Performance Linpack benchmark -- September 10, 2008 Written by A. Petitet and R. Clint Whaley, Innovative Computing Laboratory, UTK Modified by Piotr Luszczek, Innovative Computing Laboratory, UTK Modified by Julien Langou, University of Colorado Denver ================================================================================ An explanation of the input/output parameters follows: T/V : Wall time / encoded variant. N : The order of the coefficient matrix A. NB : The partitioning blocking factor. P : The number of process rows. Q : The number of process columns. Time : Time in seconds to solve the linear system. Gflops : Rate of execution for solving the linear system. The following parameter values will be used: N : 6000 NB : 960 PMAP : Row-major process mapping P : 2 Q : 2 PFACT : Left NBMIN : 2 NDIV : 4 RFACT : Left BCAST : 2ringM DEPTH : 0 SWAP : Binary-exchange L1 : no-transposed form U : no-transposed form EQUIL : no ALIGN : 8 double precision words -------------------------------------------------------------------------------- - The matrix A is randomly generated for each test. - The following scaled residual check will be computed: ||Ax-b||_oo / ( eps * ( || x ||_oo * || A ||_oo + || b ||_oo ) * N ) - The relative machine precision (eps) is taken to be 1.110223e-16 - Computational tests pass if scaled residuals are less than 16.0 Assigning device 0 to process on node tesla2.localdomain rank 0 Assigning device 3 to process on node tesla2.localdomain rank 3 Assigning device 1 to process on node tesla2.localdomain rank 1 Assigning device 2 to process on node tesla2.localdomain rank 2 DTRSM split from environment variable 0.800000 DTRSM split from environment variable 0.800000 DGEMM split from environment variable 0.900000 DGEMM split from environment variable 0.900000 DTRSM split from environment variable 0.800000 DGEMM split from environment variable 0.900000 DTRSM split from environment variable 0.800000 DGEMM split from environment variable 0.900000 ================================================================================ T/V N NB P Q Time Gflops -------------------------------------------------------------------------------- WR03L4L2 6000 960 2 2 2.84 5.078e+01 -------------------------------------------------------------------------------- ||Ax-b||_oo/(eps*(||A||_oo*||x||_oo+||b||_oo)*N)= 0.0047660 ...... PASSED ================================================================================ Finished 1 tests with the following results: 1 tests completed and passed residual checks, 0 tests completed and failed residual checks, 0 tests skipped because of illegal input values. -------------------------------------------------------------------------------- End of Tests. ================================================================================ ======================== JOB MAP ======================== Data for node: Name: tesla2 Num procs: 4 Process OMPI jobid: [48283,1] Process rank: 0 Process OMPI jobid: [48283,1] Process rank: 1 Process OMPI jobid: [48283,1] Process rank: 2 Process OMPI jobid: [48283,1] Process rank: 3 ============================================================= ================================================================================ HPLinpack 2.0 -- High-Performance Linpack benchmark -- September 10, 2008 Written by A. Petitet and R. Clint Whaley, Innovative Computing Laboratory, UTK Modified by Piotr Luszczek, Innovative Computing Laboratory, UTK Modified by Julien Langou, University of Colorado Denver ================================================================================ An explanation of the input/output parameters follows: T/V : Wall time / encoded variant. N : The order of the coefficient matrix A. NB : The partitioning blocking factor. P : The number of process rows. Q : The number of process columns. Time : Time in seconds to solve the linear system. Gflops : Rate of execution for solving the linear system. The following parameter values will be used: N : 6000 NB : 960 PMAP : Row-major process mapping P : 2 Q : 2 PFACT : Left NBMIN : 2 NDIV : 4 RFACT : Left BCAST : 2ringM DEPTH : 0 SWAP : Binary-exchange L1 : no-transposed form U : no-transposed form EQUIL : no ALIGN : 8 double precision words -------------------------------------------------------------------------------- - The matrix A is randomly generated for each test. - The following scaled residual check will be computed: ||Ax-b||_oo / ( eps * ( || x ||_oo * || A ||_oo + || b ||_oo ) * N ) - The relative machine precision (eps) is taken to be 1.110223e-16 - Computational tests pass if scaled residuals are less than 16.0 Assigning device 0 to process on node tesla2.localdomain rank 0 Assigning device 2 to process on node tesla2.localdomain rank 2 Assigning device 1 to process on node tesla2.localdomain rank 1 Assigning device 3 to process on node tesla2.localdomain rank 3 DTRSM split from environment variable 0.850000 DTRSM split from environment variable 0.850000 DGEMM split from environment variable 0.900000 DGEMM split from environment variable 0.900000 DTRSM split from environment variable 0.850000 DGEMM split from environment variable 0.900000 DTRSM split from environment variable 0.850000 DGEMM split from environment variable 0.900000 ================================================================================ T/V N NB P Q Time Gflops -------------------------------------------------------------------------------- WR03L4L2 6000 960 2 2 2.84 5.068e+01 -------------------------------------------------------------------------------- ||Ax-b||_oo/(eps*(||A||_oo*||x||_oo+||b||_oo)*N)= 0.0048664 ...... PASSED ================================================================================ Finished 1 tests with the following results: 1 tests completed and passed residual checks, 0 tests completed and failed residual checks, 0 tests skipped because of illegal input values. -------------------------------------------------------------------------------- End of Tests. ================================================================================ ======================== JOB MAP ======================== Data for node: Name: tesla2 Num procs: 4 Process OMPI jobid: [48264,1] Process rank: 0 Process OMPI jobid: [48264,1] Process rank: 1 Process OMPI jobid: [48264,1] Process rank: 2 Process OMPI jobid: [48264,1] Process rank: 3 ============================================================= ================================================================================ HPLinpack 2.0 -- High-Performance Linpack benchmark -- September 10, 2008 Written by A. Petitet and R. Clint Whaley, Innovative Computing Laboratory, UTK Modified by Piotr Luszczek, Innovative Computing Laboratory, UTK Modified by Julien Langou, University of Colorado Denver ================================================================================ An explanation of the input/output parameters follows: T/V : Wall time / encoded variant. N : The order of the coefficient matrix A. NB : The partitioning blocking factor. P : The number of process rows. Q : The number of process columns. Time : Time in seconds to solve the linear system. Gflops : Rate of execution for solving the linear system. The following parameter values will be used: N : 16000 NB : 960 PMAP : Row-major process mapping P : 2 Q : 2 PFACT : Left NBMIN : 2 NDIV : 4 RFACT : Left BCAST : 2ringM DEPTH : 0 SWAP : Binary-exchange L1 : no-transposed form U : no-transposed form EQUIL : no ALIGN : 8 double precision words -------------------------------------------------------------------------------- - The matrix A is randomly generated for each test. - The following scaled residual check will be computed: ||Ax-b||_oo / ( eps * ( || x ||_oo * || A ||_oo + || b ||_oo ) * N ) - The relative machine precision (eps) is taken to be 1.110223e-16 - Computational tests pass if scaled residuals are less than 16.0 Assigning device 0 to process on node tesla2.localdomain rank 0 Assigning device 1 to process on node tesla2.localdomain rank 1 Assigning device 2 to process on node tesla2.localdomain rank 2 Assigning device 3 to process on node tesla2.localdomain rank 3 DTRSM split from environment variable 0.750000 DTRSM split from environment variable 0.750000 DGEMM split from environment variable 0.900000 DGEMM split from environment variable 0.900000 DTRSM split from environment variable 0.750000 DGEMM split from environment variable 0.900000 DTRSM split from environment variable 0.750000 DGEMM split from environment variable 0.900000 ================================================================================ T/V N NB P Q Time Gflops -------------------------------------------------------------------------------- WR03L4L2 16000 960 2 2 23.90 1.143e+02 -------------------------------------------------------------------------------- ||Ax-b||_oo/(eps*(||A||_oo*||x||_oo+||b||_oo)*N)= 0.0055664 ...... PASSED ================================================================================ Finished 1 tests with the following results: 1 tests completed and passed residual checks, 0 tests completed and failed residual checks, 0 tests skipped because of illegal input values. -------------------------------------------------------------------------------- End of Tests. ================================================================================ ======================== JOB MAP ======================== Data for node: Name: tesla2 Num procs: 4 Process OMPI jobid: [48259,1] Process rank: 0 Process OMPI jobid: [48259,1] Process rank: 1 Process OMPI jobid: [48259,1] Process rank: 2 Process OMPI jobid: [48259,1] Process rank: 3 ============================================================= ================================================================================ HPLinpack 2.0 -- High-Performance Linpack benchmark -- September 10, 2008 Written by A. Petitet and R. Clint Whaley, Innovative Computing Laboratory, UTK Modified by Piotr Luszczek, Innovative Computing Laboratory, UTK Modified by Julien Langou, University of Colorado Denver ================================================================================ An explanation of the input/output parameters follows: T/V : Wall time / encoded variant. N : The order of the coefficient matrix A. NB : The partitioning blocking factor. P : The number of process rows. Q : The number of process columns. Time : Time in seconds to solve the linear system. Gflops : Rate of execution for solving the linear system. The following parameter values will be used: N : 26000 NB : 960 PMAP : Row-major process mapping P : 2 Q : 2 PFACT : Left NBMIN : 2 NDIV : 4 RFACT : Left BCAST : 2ringM DEPTH : 0 SWAP : Binary-exchange L1 : no-transposed form U : no-transposed form EQUIL : no ALIGN : 8 double precision words -------------------------------------------------------------------------------- - The matrix A is randomly generated for each test. - The following scaled residual check will be computed: ||Ax-b||_oo / ( eps * ( || x ||_oo * || A ||_oo + || b ||_oo ) * N ) - The relative machine precision (eps) is taken to be 1.110223e-16 - Computational tests pass if scaled residuals are less than 16.0 Assigning device 0 to process on node tesla2.localdomain rank 0 Assigning device 3 to process on node tesla2.localdomain rank 3 Assigning device 1 to process on node tesla2.localdomain rank 1 Assigning device 2 to process on node tesla2.localdomain rank 2 DTRSM split from environment variable 0.750000 DTRSM split from environment variable 0.750000 DGEMM split from environment variable 0.900000 DGEMM split from environment variable 0.900000 DTRSM split from environment variable 0.750000 DGEMM split from environment variable 0.900000 DTRSM split from environment variable 0.750000 DGEMM split from environment variable 0.900000 ================================================================================ T/V N NB P Q Time Gflops -------------------------------------------------------------------------------- WR03L4L2 26000 960 2 2 73.38 1.597e+02 -------------------------------------------------------------------------------- ||Ax-b||_oo/(eps*(||A||_oo*||x||_oo+||b||_oo)*N)= 0.0040649 ...... PASSED ================================================================================ Finished 1 tests with the following results: 1 tests completed and passed residual checks, 0 tests completed and failed residual checks, 0 tests skipped because of illegal input values. -------------------------------------------------------------------------------- End of Tests. ================================================================================ ======================== JOB MAP ======================== Data for node: Name: tesla2 Num procs: 4 Process OMPI jobid: [48471,1] Process rank: 0 Process OMPI jobid: [48471,1] Process rank: 1 Process OMPI jobid: [48471,1] Process rank: 2 Process OMPI jobid: [48471,1] Process rank: 3 ============================================================= ================================================================================ HPLinpack 2.0 -- High-Performance Linpack benchmark -- September 10, 2008 Written by A. Petitet and R. Clint Whaley, Innovative Computing Laboratory, UTK Modified by Piotr Luszczek, Innovative Computing Laboratory, UTK Modified by Julien Langou, University of Colorado Denver ================================================================================ An explanation of the input/output parameters follows: T/V : Wall time / encoded variant. N : The order of the coefficient matrix A. NB : The partitioning blocking factor. P : The number of process rows. Q : The number of process columns. Time : Time in seconds to solve the linear system. Gflops : Rate of execution for solving the linear system. The following parameter values will be used: N : 36000 NB : 960 PMAP : Row-major process mapping P : 2 Q : 2 PFACT : Left NBMIN : 2 NDIV : 4 RFACT : Left BCAST : 2ringM DEPTH : 0 SWAP : Binary-exchange L1 : no-transposed form U : no-transposed form EQUIL : no ALIGN : 8 double precision words -------------------------------------------------------------------------------- - The matrix A is randomly generated for each test. - The following scaled residual check will be computed: ||Ax-b||_oo / ( eps * ( || x ||_oo * || A ||_oo + || b ||_oo ) * N ) - The relative machine precision (eps) is taken to be 1.110223e-16 - Computational tests pass if scaled residuals are less than 16.0 Assigning device 0 to process on node tesla2.localdomain rank 0 Assigning device 1 to process on node tesla2.localdomain rank 1 Assigning device 2 to process on node tesla2.localdomain rank 2 Assigning device 3 to process on node tesla2.localdomain rank 3 DTRSM split from environment variable 0.750000 DGEMM split from environment variable 0.900000 DTRSM split from environment variable 0.750000 DGEMM split from environment variable 0.900000 DTRSM split from environment variable 0.750000 DGEMM split from environment variable 0.900000 DTRSM split from environment variable 0.750000 DGEMM split from environment variable 0.900000 ================================================================================ T/V N NB P Q Time Gflops -------------------------------------------------------------------------------- WR03L4L2 36000 960 2 2 166.09 1.873e+02 -------------------------------------------------------------------------------- ||Ax-b||_oo/(eps*(||A||_oo*||x||_oo+||b||_oo)*N)= 0.0041700 ...... PASSED ================================================================================ Finished 1 tests with the following results: 1 tests completed and passed residual checks, 0 tests completed and failed residual checks, 0 tests skipped because of illegal input values. -------------------------------------------------------------------------------- End of Tests. ================================================================================ ======================== JOB MAP ======================== Data for node: Name: tesla2 Num procs: 4 Process OMPI jobid: [48455,1] Process rank: 0 Process OMPI jobid: [48455,1] Process rank: 1 Process OMPI jobid: [48455,1] Process rank: 2 Process OMPI jobid: [48455,1] Process rank: 3 ============================================================= ================================================================================ HPLinpack 2.0 -- High-Performance Linpack benchmark -- September 10, 2008 Written by A. Petitet and R. Clint Whaley, Innovative Computing Laboratory, UTK Modified by Piotr Luszczek, Innovative Computing Laboratory, UTK Modified by Julien Langou, University of Colorado Denver ================================================================================ An explanation of the input/output parameters follows: T/V : Wall time / encoded variant. N : The order of the coefficient matrix A. NB : The partitioning blocking factor. P : The number of process rows. Q : The number of process columns. Time : Time in seconds to solve the linear system. Gflops : Rate of execution for solving the linear system. The following parameter values will be used: N : 46000 NB : 960 PMAP : Row-major process mapping P : 2 Q : 2 PFACT : Left NBMIN : 2 NDIV : 4 RFACT : Left BCAST : 2ringM DEPTH : 0 SWAP : Binary-exchange L1 : no-transposed form U : no-transposed form EQUIL : no ALIGN : 8 double precision words -------------------------------------------------------------------------------- - The matrix A is randomly generated for each test. - The following scaled residual check will be computed: ||Ax-b||_oo / ( eps * ( || x ||_oo * || A ||_oo + || b ||_oo ) * N ) - The relative machine precision (eps) is taken to be 1.110223e-16 - Computational tests pass if scaled residuals are less than 16.0 Assigning device 0 to process on node tesla2.localdomain rank 0 Assigning device 1 to process on node tesla2.localdomain rank 1 Assigning device 2 to process on node tesla2.localdomain rank 2 Assigning device 3 to process on node tesla2.localdomain rank 3 DTRSM split from environment variable 0.750000 DTRSM split from environment variable 0.750000 DGEMM split from environment variable 0.900000 DGEMM split from environment variable 0.900000 DTRSM split from environment variable 0.750000 DGEMM split from environment variable 0.900000 DTRSM split from environment variable 0.750000 DGEMM split from environment variable 0.900000 ================================================================================ T/V N NB P Q Time Gflops -------------------------------------------------------------------------------- WR03L4L2 46000 960 2 2 328.96 1.973e+02 -------------------------------------------------------------------------------- ||Ax-b||_oo/(eps*(||A||_oo*||x||_oo+||b||_oo)*N)= 0.0042671 ...... PASSED ================================================================================ Finished 1 tests with the following results: 1 tests completed and passed residual checks, 0 tests completed and failed residual checks, 0 tests skipped because of illegal input values. -------------------------------------------------------------------------------- End of Tests. ================================================================================ ======================== JOB MAP ======================== Data for node: Name: tesla2 Num procs: 4 Process OMPI jobid: [48425,1] Process rank: 0 Process OMPI jobid: [48425,1] Process rank: 1 Process OMPI jobid: [48425,1] Process rank: 2 Process OMPI jobid: [48425,1] Process rank: 3 ============================================================= ================================================================================ HPLinpack 2.0 -- High-Performance Linpack benchmark -- September 10, 2008 Written by A. Petitet and R. Clint Whaley, Innovative Computing Laboratory, UTK Modified by Piotr Luszczek, Innovative Computing Laboratory, UTK Modified by Julien Langou, University of Colorado Denver ================================================================================ An explanation of the input/output parameters follows: T/V : Wall time / encoded variant. N : The order of the coefficient matrix A. NB : The partitioning blocking factor. P : The number of process rows. Q : The number of process columns. Time : Time in seconds to solve the linear system. Gflops : Rate of execution for solving the linear system. The following parameter values will be used: N : 56000 NB : 960 PMAP : Row-major process mapping P : 2 Q : 2 PFACT : Left NBMIN : 2 NDIV : 4 RFACT : Left BCAST : 2ringM DEPTH : 0 SWAP : Binary-exchange L1 : no-transposed form U : no-transposed form EQUIL : no ALIGN : 8 double precision words -------------------------------------------------------------------------------- - The matrix A is randomly generated for each test. - The following scaled residual check will be computed: ||Ax-b||_oo / ( eps * ( || x ||_oo * || A ||_oo + || b ||_oo ) * N ) - The relative machine precision (eps) is taken to be 1.110223e-16 - Computational tests pass if scaled residuals are less than 16.0 Assigning device 0 to process on node tesla2.localdomain rank 0 Assigning device 1 to process on node tesla2.localdomain rank 1 Assigning device 2 to process on node tesla2.localdomain rank 2 Assigning device 3 to process on node tesla2.localdomain rank 3 DTRSM split from environment variable 0.750000 DTRSM split from environment variable 0.750000 DGEMM split from environment variable 0.900000 DGEMM split from environment variable 0.900000 DTRSM split from environment variable 0.750000 DGEMM split from environment variable 0.900000 DTRSM split from environment variable 0.750000 DGEMM split from environment variable 0.900000 ================================================================================ T/V N NB P Q Time Gflops -------------------------------------------------------------------------------- WR03L4L2 56000 960 2 2 555.08 2.109e+02 -------------------------------------------------------------------------------- ||Ax-b||_oo/(eps*(||A||_oo*||x||_oo+||b||_oo)*N)= 0.0039375 ...... PASSED ================================================================================ Finished 1 tests with the following results: 1 tests completed and passed residual checks, 0 tests completed and failed residual checks, 0 tests skipped because of illegal input values. -------------------------------------------------------------------------------- End of Tests. ================================================================================ ======================== JOB MAP ======================== Data for node: Name: tesla2 Num procs: 4 Process OMPI jobid: [48399,1] Process rank: 0 Process OMPI jobid: [48399,1] Process rank: 1 Process OMPI jobid: [48399,1] Process rank: 2 Process OMPI jobid: [48399,1] Process rank: 3 ============================================================= ================================================================================ HPLinpack 2.0 -- High-Performance Linpack benchmark -- September 10, 2008 Written by A. Petitet and R. Clint Whaley, Innovative Computing Laboratory, UTK Modified by Piotr Luszczek, Innovative Computing Laboratory, UTK Modified by Julien Langou, University of Colorado Denver ================================================================================ An explanation of the input/output parameters follows: T/V : Wall time / encoded variant. N : The order of the coefficient matrix A. NB : The partitioning blocking factor. P : The number of process rows. Q : The number of process columns. Time : Time in seconds to solve the linear system. Gflops : Rate of execution for solving the linear system. The following parameter values will be used: N : 56000 NB : 1280 PMAP : Row-major process mapping P : 2 Q : 2 PFACT : Left NBMIN : 2 NDIV : 4 RFACT : Left BCAST : 2ringM DEPTH : 0 SWAP : Binary-exchange L1 : no-transposed form U : no-transposed form EQUIL : no ALIGN : 8 double precision words -------------------------------------------------------------------------------- - The matrix A is randomly generated for each test. - The following scaled residual check will be computed: ||Ax-b||_oo / ( eps * ( || x ||_oo * || A ||_oo + || b ||_oo ) * N ) - The relative machine precision (eps) is taken to be 1.110223e-16 - Computational tests pass if scaled residuals are less than 16.0 Assigning device 0 to process on node tesla2.localdomain rank 0 Assigning device 1 to process on node tesla2.localdomain rank 1 Assigning device 3 to process on node tesla2.localdomain rank 3 Assigning device 2 to process on node tesla2.localdomain rank 2 DTRSM split from environment variable 0.750000 DTRSM split from environment variable 0.750000 DGEMM split from environment variable 0.900000 DGEMM split from environment variable 0.900000 DTRSM split from environment variable 0.750000 DGEMM split from environment variable 0.900000 DTRSM split from environment variable 0.750000 DGEMM split from environment variable 0.900000 ================================================================================ T/V N NB P Q Time Gflops -------------------------------------------------------------------------------- WR03L4L2 56000 1280 2 2 556.00 2.106e+02 -------------------------------------------------------------------------------- ||Ax-b||_oo/(eps*(||A||_oo*||x||_oo+||b||_oo)*N)= 0.0038182 ...... PASSED ================================================================================ Finished 1 tests with the following results: 1 tests completed and passed residual checks, 0 tests completed and failed residual checks, 0 tests skipped because of illegal input values. -------------------------------------------------------------------------------- End of Tests. ================================================================================ ======================== JOB MAP ======================== Data for node: Name: tesla2 Num procs: 4 Process OMPI jobid: [48621,1] Process rank: 0 Process OMPI jobid: [48621,1] Process rank: 1 Process OMPI jobid: [48621,1] Process rank: 2 Process OMPI jobid: [48621,1] Process rank: 3 ============================================================= ================================================================================ HPLinpack 2.0 -- High-Performance Linpack benchmark -- September 10, 2008 Written by A. Petitet and R. Clint Whaley, Innovative Computing Laboratory, UTK Modified by Piotr Luszczek, Innovative Computing Laboratory, UTK Modified by Julien Langou, University of Colorado Denver ================================================================================ An explanation of the input/output parameters follows: T/V : Wall time / encoded variant. N : The order of the coefficient matrix A. NB : The partitioning blocking factor. P : The number of process rows. Q : The number of process columns. Time : Time in seconds to solve the linear system. Gflops : Rate of execution for solving the linear system. The following parameter values will be used: N : 56000 NB : 1120 PMAP : Row-major process mapping P : 2 Q : 2 PFACT : Left NBMIN : 2 NDIV : 4 RFACT : Left BCAST : 2ringM DEPTH : 0 SWAP : Binary-exchange L1 : no-transposed form U : no-transposed form EQUIL : no ALIGN : 8 double precision words -------------------------------------------------------------------------------- - The matrix A is randomly generated for each test. - The following scaled residual check will be computed: ||Ax-b||_oo / ( eps * ( || x ||_oo * || A ||_oo + || b ||_oo ) * N ) - The relative machine precision (eps) is taken to be 1.110223e-16 - Computational tests pass if scaled residuals are less than 16.0 Assigning device 0 to process on node tesla2.localdomain rank 0 Assigning device 1 to process on node tesla2.localdomain rank 1 Assigning device 3 to process on node tesla2.localdomain rank 3 Assigning device 2 to process on node tesla2.localdomain rank 2 DTRSM split from environment variable 0.750000 DTRSM split from environment variable 0.750000 DGEMM split from environment variable 0.900000 DGEMM split from environment variable 0.900000 DTRSM split from environment variable 0.750000 DGEMM split from environment variable 0.900000 DTRSM split from environment variable 0.750000 DGEMM split from environment variable 0.900000 ================================================================================ T/V N NB P Q Time Gflops -------------------------------------------------------------------------------- WR03L4L2 56000 1120 2 2 539.51 2.170e+02 -------------------------------------------------------------------------------- ||Ax-b||_oo/(eps*(||A||_oo*||x||_oo+||b||_oo)*N)= 0.0039647 ...... PASSED ================================================================================ Finished 1 tests with the following results: 1 tests completed and passed residual checks, 0 tests completed and failed residual checks, 0 tests skipped because of illegal input values. -------------------------------------------------------------------------------- End of Tests. ================================================================================ ======================== JOB MAP ======================== Data for node: Name: tesla2 Num procs: 4 Process OMPI jobid: [48585,1] Process rank: 0 Process OMPI jobid: [48585,1] Process rank: 1 Process OMPI jobid: [48585,1] Process rank: 2 Process OMPI jobid: [48585,1] Process rank: 3 ============================================================= ================================================================================ HPLinpack 2.0 -- High-Performance Linpack benchmark -- September 10, 2008 Written by A. Petitet and R. Clint Whaley, Innovative Computing Laboratory, UTK Modified by Piotr Luszczek, Innovative Computing Laboratory, UTK Modified by Julien Langou, University of Colorado Denver ================================================================================ An explanation of the input/output parameters follows: T/V : Wall time / encoded variant. N : The order of the coefficient matrix A. NB : The partitioning blocking factor. P : The number of process rows. Q : The number of process columns. Time : Time in seconds to solve the linear system. Gflops : Rate of execution for solving the linear system. The following parameter values will be used: N : 56000 NB : 1120 PMAP : Row-major process mapping P : 2 Q : 2 PFACT : Left NBMIN : 2 NDIV : 4 RFACT : Left BCAST : 1ring DEPTH : 0 SWAP : Binary-exchange L1 : no-transposed form U : no-transposed form EQUIL : no ALIGN : 8 double precision words -------------------------------------------------------------------------------- - The matrix A is randomly generated for each test. - The following scaled residual check will be computed: ||Ax-b||_oo / ( eps * ( || x ||_oo * || A ||_oo + || b ||_oo ) * N ) - The relative machine precision (eps) is taken to be 1.110223e-16 - Computational tests pass if scaled residuals are less than 16.0 Assigning device 0 to process on node tesla2.localdomain rank 0 Assigning device 1 to process on node tesla2.localdomain rank 1 Assigning device 2 to process on node tesla2.localdomain rank 2 Assigning device 3 to process on node tesla2.localdomain rank 3 ======================== JOB MAP ======================== Data for node: Name: tesla2 Num procs: 4 Process OMPI jobid: [48573,1] Process rank: 0 Process OMPI jobid: [48573,1] Process rank: 1 Process OMPI jobid: [48573,1] Process rank: 2 Process OMPI jobid: [48573,1] Process rank: 3 ============================================================= ================================================================================ HPLinpack 2.0 -- High-Performance Linpack benchmark -- September 10, 2008 Written by A. Petitet and R. Clint Whaley, Innovative Computing Laboratory, UTK Modified by Piotr Luszczek, Innovative Computing Laboratory, UTK Modified by Julien Langou, University of Colorado Denver ================================================================================ An explanation of the input/output parameters follows: T/V : Wall time / encoded variant. N : The order of the coefficient matrix A. NB : The partitioning blocking factor. P : The number of process rows. Q : The number of process columns. Time : Time in seconds to solve the linear system. Gflops : Rate of execution for solving the linear system. The following parameter values will be used: N : 56000 NB : 1120 PMAP : Row-major process mapping P : 2 Q : 2 PFACT : Left NBMIN : 2 NDIV : 4 RFACT : Left BCAST : 1ring DEPTH : 0 SWAP : Binary-exchange L1 : no-transposed form U : no-transposed form EQUIL : no ALIGN : 8 double precision words -------------------------------------------------------------------------------- - The matrix A is randomly generated for each test. - The following scaled residual check will be computed: ||Ax-b||_oo / ( eps * ( || x ||_oo * || A ||_oo + || b ||_oo ) * N ) - The relative machine precision (eps) is taken to be 1.110223e-16 - Computational tests pass if scaled residuals are less than 16.0 Assigning device 1 to process on node tesla2.localdomain rank 1 Assigning device 2 to process on node tesla2.localdomain rank 2 Assigning device 3 to process on node tesla2.localdomain rank 3 Assigning device 0 to process on node tesla2.localdomain rank 0 DTRSM split from environment variable 0.750000 DTRSM split from environment variable 0.750000 DGEMM split from environment variable 0.900000 DGEMM split from environment variable 0.900000 DTRSM split from environment variable 0.750000 DGEMM split from environment variable 0.900000 DTRSM split from environment variable 0.750000 DGEMM split from environment variable 0.900000 ================================================================================ T/V N NB P Q Time Gflops -------------------------------------------------------------------------------- WR00L4L2 56000 1120 2 2 539.57 2.170e+02 -------------------------------------------------------------------------------- ||Ax-b||_oo/(eps*(||A||_oo*||x||_oo+||b||_oo)*N)= 0.0039647 ...... PASSED ================================================================================ Finished 1 tests with the following results: 1 tests completed and passed residual checks, 0 tests completed and failed residual checks, 0 tests skipped because of illegal input values. -------------------------------------------------------------------------------- End of Tests. ================================================================================ ======================== JOB MAP ======================== Data for node: Name: tesla2 Num procs: 4 Process OMPI jobid: [48541,1] Process rank: 0 Process OMPI jobid: [48541,1] Process rank: 1 Process OMPI jobid: [48541,1] Process rank: 2 Process OMPI jobid: [48541,1] Process rank: 3 ============================================================= ================================================================================ HPLinpack 2.0 -- High-Performance Linpack benchmark -- September 10, 2008 Written by A. Petitet and R. Clint Whaley, Innovative Computing Laboratory, UTK Modified by Piotr Luszczek, Innovative Computing Laboratory, UTK Modified by Julien Langou, University of Colorado Denver ================================================================================ An explanation of the input/output parameters follows: T/V : Wall time / encoded variant. N : The order of the coefficient matrix A. NB : The partitioning blocking factor. P : The number of process rows. Q : The number of process columns. Time : Time in seconds to solve the linear system. Gflops : Rate of execution for solving the linear system. The following parameter values will be used: N : 56000 NB : 1180 PMAP : Row-major process mapping P : 2 Q : 2 PFACT : Left NBMIN : 2 NDIV : 4 RFACT : Left BCAST : 1ring DEPTH : 0 SWAP : Binary-exchange L1 : no-transposed form U : no-transposed form EQUIL : no ALIGN : 8 double precision words -------------------------------------------------------------------------------- - The matrix A is randomly generated for each test. - The following scaled residual check will be computed: ||Ax-b||_oo / ( eps * ( || x ||_oo * || A ||_oo + || b ||_oo ) * N ) - The relative machine precision (eps) is taken to be 1.110223e-16 - Computational tests pass if scaled residuals are less than 16.0 Assigning device 0 to process on node tesla2.localdomain rank 0 Assigning device 1 to process on node tesla2.localdomain rank 1 Assigning device 2 to process on node tesla2.localdomain rank 2 Assigning device 3 to process on node tesla2.localdomain rank 3 DTRSM split from environment variable 0.750000 DTRSM split from environment variable 0.750000 DGEMM split from environment variable 0.900000 DGEMM split from environment variable 0.900000 DTRSM split from environment variable 0.750000 DGEMM split from environment variable 0.900000 DTRSM split from environment variable 0.750000 DGEMM split from environment variable 0.900000 ================================================================================ T/V N NB P Q Time Gflops -------------------------------------------------------------------------------- WR00L4L2 56000 1180 2 2 666.46 1.757e+02 -------------------------------------------------------------------------------- ||Ax-b||_oo/(eps*(||A||_oo*||x||_oo+||b||_oo)*N)= 0.0041903 ...... PASSED ================================================================================ Finished 1 tests with the following results: 1 tests completed and passed residual checks, 0 tests completed and failed residual checks, 0 tests skipped because of illegal input values. -------------------------------------------------------------------------------- End of Tests. ================================================================================ ======================== JOB MAP ======================== Data for node: Name: tesla2.localdomain Num procs: 4 Process OMPI jobid: [45659,1] Process rank: 0 Process OMPI jobid: [45659,1] Process rank: 1 Process OMPI jobid: [45659,1] Process rank: 2 Process OMPI jobid: [45659,1] Process rank: 3 ============================================================= ================================================================================ HPLinpack 2.0 -- High-Performance Linpack benchmark -- September 10, 2008 Written by A. Petitet and R. Clint Whaley, Innovative Computing Laboratory, UTK Modified by Piotr Luszczek, Innovative Computing Laboratory, UTK Modified by Julien Langou, University of Colorado Denver ================================================================================ An explanation of the input/output parameters follows: T/V : Wall time / encoded variant. N : The order of the coefficient matrix A. NB : The partitioning blocking factor. P : The number of process rows. Q : The number of process columns. Time : Time in seconds to solve the linear system. Gflops : Rate of execution for solving the linear system. The following parameter values will be used: N : 16000 NB : 1140 PMAP : Row-major process mapping P : 2 Q : 2 PFACT : Left NBMIN : 2 NDIV : 4 RFACT : Left BCAST : 1ring DEPTH : 0 SWAP : Binary-exchange L1 : no-transposed form U : no-transposed form EQUIL : no ALIGN : 8 double precision words -------------------------------------------------------------------------------- - The matrix A is randomly generated for each test. - The following scaled residual check will be computed: ||Ax-b||_oo / ( eps * ( || x ||_oo * || A ||_oo + || b ||_oo ) * N ) - The relative machine precision (eps) is taken to be 1.110223e-16 - Computational tests pass if scaled residuals are less than 16.0 Assigning device 0 to process on node tesla2.localdomain rank 0 Assigning device 2 to process on node tesla2.localdomain rank 2 Assigning device 3 to process on node tesla2.localdomain rank 3 Assigning device 1 to process on node tesla2.localdomain rank 1 DTRSM split from environment variable 0.750000 DTRSM split from environment variable 0.750000 DGEMM split from environment variable 0.900000 DGEMM split from environment variable 0.900000 DTRSM split from environment variable 0.750000 DGEMM split from environment variable 0.900000 DTRSM split from environment variable 0.750000 DGEMM split from environment variable 0.900000 ================================================================================ T/V N NB P Q Time Gflops -------------------------------------------------------------------------------- WR00L4L2 16000 1140 2 2 26.63 1.026e+02 -------------------------------------------------------------------------------- ||Ax-b||_oo/(eps*(||A||_oo*||x||_oo+||b||_oo)*N)= 0.0061187 ...... PASSED ================================================================================ Finished 1 tests with the following results: 1 tests completed and passed residual checks, 0 tests completed and failed residual checks, 0 tests skipped because of illegal input values. -------------------------------------------------------------------------------- End of Tests. ================================================================================ ======================== JOB MAP ======================== Data for node: Name: tesla2.localdomain Num procs: 4 Process OMPI jobid: [45684,1] Process rank: 0 Process OMPI jobid: [45684,1] Process rank: 1 Process OMPI jobid: [45684,1] Process rank: 2 Process OMPI jobid: [45684,1] Process rank: 3 ============================================================= ================================================================================ HPLinpack 2.0 -- High-Performance Linpack benchmark -- September 10, 2008 Written by A. Petitet and R. Clint Whaley, Innovative Computing Laboratory, UTK Modified by Piotr Luszczek, Innovative Computing Laboratory, UTK Modified by Julien Langou, University of Colorado Denver ================================================================================ An explanation of the input/output parameters follows: T/V : Wall time / encoded variant. N : The order of the coefficient matrix A. NB : The partitioning blocking factor. P : The number of process rows. Q : The number of process columns. Time : Time in seconds to solve the linear system. Gflops : Rate of execution for solving the linear system. The following parameter values will be used: N : 16000 NB : 1120 PMAP : Row-major process mapping P : 2 Q : 2 PFACT : Left NBMIN : 2 NDIV : 4 RFACT : Left BCAST : 1ring DEPTH : 0 SWAP : Binary-exchange L1 : no-transposed form U : no-transposed form EQUIL : no ALIGN : 8 double precision words -------------------------------------------------------------------------------- - The matrix A is randomly generated for each test. - The following scaled residual check will be computed: ||Ax-b||_oo / ( eps * ( || x ||_oo * || A ||_oo + || b ||_oo ) * N ) - The relative machine precision (eps) is taken to be 1.110223e-16 - Computational tests pass if scaled residuals are less than 16.0 Assigning device 0 to process on node tesla2.localdomain rank 0 Assigning device 1 to process on node tesla2.localdomain rank 1 Assigning device 2 to process on node tesla2.localdomain rank 2 Assigning device 3 to process on node tesla2.localdomain rank 3 DTRSM split from environment variable 0.750000 DTRSM split from environment variable 0.750000 DGEMM split from environment variable 0.900000 DGEMM split from environment variable 0.900000 DTRSM split from environment variable 0.750000 DGEMM split from environment variable 0.900000 DTRSM split from environment variable 0.750000 DGEMM split from environment variable 0.900000 ================================================================================ T/V N NB P Q Time Gflops -------------------------------------------------------------------------------- WR00L4L2 16000 1120 2 2 22.72 1.202e+02 -------------------------------------------------------------------------------- ||Ax-b||_oo/(eps*(||A||_oo*||x||_oo+||b||_oo)*N)= 0.0048042 ...... PASSED ================================================================================ Finished 1 tests with the following results: 1 tests completed and passed residual checks, 0 tests completed and failed residual checks, 0 tests skipped because of illegal input values. -------------------------------------------------------------------------------- End of Tests. ================================================================================ ======================== JOB MAP ======================== Data for node: Name: tesla2.localdomain Num procs: 4 Process OMPI jobid: [45579,1] Process rank: 0 Process OMPI jobid: [45579,1] Process rank: 1 Process OMPI jobid: [45579,1] Process rank: 2 Process OMPI jobid: [45579,1] Process rank: 3 ============================================================= ================================================================================ HPLinpack 2.0 -- High-Performance Linpack benchmark -- September 10, 2008 Written by A. Petitet and R. Clint Whaley, Innovative Computing Laboratory, UTK Modified by Piotr Luszczek, Innovative Computing Laboratory, UTK Modified by Julien Langou, University of Colorado Denver ================================================================================ An explanation of the input/output parameters follows: T/V : Wall time / encoded variant. N : The order of the coefficient matrix A. NB : The partitioning blocking factor. P : The number of process rows. Q : The number of process columns. Time : Time in seconds to solve the linear system. Gflops : Rate of execution for solving the linear system. The following parameter values will be used: N : 16000 NB : 1100 PMAP : Row-major process mapping P : 2 Q : 2 PFACT : Left NBMIN : 2 NDIV : 4 RFACT : Left BCAST : 1ring DEPTH : 0 SWAP : Binary-exchange L1 : no-transposed form U : no-transposed form EQUIL : no ALIGN : 8 double precision words -------------------------------------------------------------------------------- - The matrix A is randomly generated for each test. - The following scaled residual check will be computed: ||Ax-b||_oo / ( eps * ( || x ||_oo * || A ||_oo + || b ||_oo ) * N ) - The relative machine precision (eps) is taken to be 1.110223e-16 - Computational tests pass if scaled residuals are less than 16.0 Assigning device 0 to process on node tesla2.localdomain rank 0 Assigning device 1 to process on node tesla2.localdomain rank 1 Assigning device 2 to process on node tesla2.localdomain rank 2 Assigning device 3 to process on node tesla2.localdomain rank 3 DTRSM split from environment variable 0.750000 DTRSM split from environment variable 0.750000 DGEMM split from environment variable 0.900000 DGEMM split from environment variable 0.900000 DTRSM split from environment variable 0.750000 DGEMM split from environment variable 0.900000 DTRSM split from environment variable 0.750000 DGEMM split from environment variable 0.900000 ================================================================================ T/V N NB P Q Time Gflops -------------------------------------------------------------------------------- WR00L4L2 16000 1100 2 2 26.01 1.050e+02 -------------------------------------------------------------------------------- ||Ax-b||_oo/(eps*(||A||_oo*||x||_oo+||b||_oo)*N)= 0.0049953 ...... PASSED ================================================================================ Finished 1 tests with the following results: 1 tests completed and passed residual checks, 0 tests completed and failed residual checks, 0 tests skipped because of illegal input values. -------------------------------------------------------------------------------- End of Tests. ================================================================================ ======================== JOB MAP ======================== Data for node: Name: tesla2.localdomain Num procs: 4 Process OMPI jobid: [45600,1] Process rank: 0 Process OMPI jobid: [45600,1] Process rank: 1 Process OMPI jobid: [45600,1] Process rank: 2 Process OMPI jobid: [45600,1] Process rank: 3 ============================================================= ================================================================================ HPLinpack 2.0 -- High-Performance Linpack benchmark -- September 10, 2008 Written by A. Petitet and R. Clint Whaley, Innovative Computing Laboratory, UTK Modified by Piotr Luszczek, Innovative Computing Laboratory, UTK Modified by Julien Langou, University of Colorado Denver ================================================================================ An explanation of the input/output parameters follows: T/V : Wall time / encoded variant. N : The order of the coefficient matrix A. NB : The partitioning blocking factor. P : The number of process rows. Q : The number of process columns. Time : Time in seconds to solve the linear system. Gflops : Rate of execution for solving the linear system. The following parameter values will be used: N : 16000 NB : 1080 PMAP : Row-major process mapping P : 2 Q : 2 PFACT : Left NBMIN : 2 NDIV : 4 RFACT : Left BCAST : 1ring DEPTH : 0 SWAP : Binary-exchange L1 : no-transposed form U : no-transposed form EQUIL : no ALIGN : 8 double precision words -------------------------------------------------------------------------------- - The matrix A is randomly generated for each test. - The following scaled residual check will be computed: ||Ax-b||_oo / ( eps * ( || x ||_oo * || A ||_oo + || b ||_oo ) * N ) - The relative machine precision (eps) is taken to be 1.110223e-16 - Computational tests pass if scaled residuals are less than 16.0 Assigning device 0 to process on node tesla2.localdomain rank 0 Assigning device 3 to process on node tesla2.localdomain rank 3 Assigning device 1 to process on node tesla2.localdomain rank 1 Assigning device 2 to process on node tesla2.localdomain rank 2 DTRSM split from environment variable 0.750000 DGEMM split from environment variable 0.900000 DTRSM split from environment variable 0.750000 DGEMM split from environment variable 0.900000 DTRSM split from environment variable 0.750000 DGEMM split from environment variable 0.900000 DTRSM split from environment variable 0.750000 DGEMM split from environment variable 0.900000 ================================================================================ T/V N NB P Q Time Gflops -------------------------------------------------------------------------------- WR00L4L2 16000 1080 2 2 25.57 1.068e+02 -------------------------------------------------------------------------------- ||Ax-b||_oo/(eps*(||A||_oo*||x||_oo+||b||_oo)*N)= 0.0042254 ...... PASSED ================================================================================ Finished 1 tests with the following results: 1 tests completed and passed residual checks, 0 tests completed and failed residual checks, 0 tests skipped because of illegal input values. -------------------------------------------------------------------------------- End of Tests. ================================================================================ ======================== JOB MAP ======================== Data for node: Name: tesla2.localdomain Num procs: 4 Process OMPI jobid: [45624,1] Process rank: 0 Process OMPI jobid: [45624,1] Process rank: 1 Process OMPI jobid: [45624,1] Process rank: 2 Process OMPI jobid: [45624,1] Process rank: 3 ============================================================= ================================================================================ HPLinpack 2.0 -- High-Performance Linpack benchmark -- September 10, 2008 Written by A. Petitet and R. Clint Whaley, Innovative Computing Laboratory, UTK Modified by Piotr Luszczek, Innovative Computing Laboratory, UTK Modified by Julien Langou, University of Colorado Denver ================================================================================ An explanation of the input/output parameters follows: T/V : Wall time / encoded variant. N : The order of the coefficient matrix A. NB : The partitioning blocking factor. P : The number of process rows. Q : The number of process columns. Time : Time in seconds to solve the linear system. Gflops : Rate of execution for solving the linear system. The following parameter values will be used: N : 16000 NB : 960 980 1000 1020 1040 1060 1120 1130 1140 PMAP : Row-major process mapping P : 2 Q : 2 PFACT : Left NBMIN : 2 NDIV : 4 RFACT : Left BCAST : 1ring DEPTH : 0 SWAP : Binary-exchange L1 : no-transposed form U : no-transposed form EQUIL : no ALIGN : 8 double precision words -------------------------------------------------------------------------------- - The matrix A is randomly generated for each test. - The following scaled residual check will be computed: ||Ax-b||_oo / ( eps * ( || x ||_oo * || A ||_oo + || b ||_oo ) * N ) - The relative machine precision (eps) is taken to be 1.110223e-16 - Computational tests pass if scaled residuals are less than 16.0 Assigning device 0 to process on node tesla2.localdomain rank 0 Assigning device 1 to process on node tesla2.localdomain rank 1 Assigning device 2 to process on node tesla2.localdomain rank 2 Assigning device 3 to process on node tesla2.localdomain rank 3 DTRSM split from environment variable 0.750000 DTRSM split from environment variable 0.750000 DGEMM split from environment variable 0.900000 DGEMM split from environment variable 0.900000 DTRSM split from environment variable 0.750000 DGEMM split from environment variable 0.900000 DTRSM split from environment variable 0.750000 DGEMM split from environment variable 0.900000 ================================================================================ T/V N NB P Q Time Gflops -------------------------------------------------------------------------------- WR00L4L2 16000 960 2 2 23.47 1.163e+02 -------------------------------------------------------------------------------- ||Ax-b||_oo/(eps*(||A||_oo*||x||_oo+||b||_oo)*N)= 0.0055664 ...... PASSED ================================================================================ T/V N NB P Q Time Gflops -------------------------------------------------------------------------------- WR00L4L2 16000 980 2 2 25.75 1.061e+02 -------------------------------------------------------------------------------- ||Ax-b||_oo/(eps*(||A||_oo*||x||_oo+||b||_oo)*N)= 0.0046083 ...... PASSED ================================================================================ T/V N NB P Q Time Gflops -------------------------------------------------------------------------------- WR00L4L2 16000 1000 2 2 25.98 1.051e+02 -------------------------------------------------------------------------------- ||Ax-b||_oo/(eps*(||A||_oo*||x||_oo+||b||_oo)*N)= 0.0046798 ...... PASSED ================================================================================ T/V N NB P Q Time Gflops -------------------------------------------------------------------------------- WR00L4L2 16000 1020 2 2 25.83 1.057e+02 -------------------------------------------------------------------------------- ||Ax-b||_oo/(eps*(||A||_oo*||x||_oo+||b||_oo)*N)= 0.0050808 ...... PASSED ================================================================================ T/V N NB P Q Time Gflops -------------------------------------------------------------------------------- WR00L4L2 16000 1040 2 2 24.17 1.130e+02 -------------------------------------------------------------------------------- ||Ax-b||_oo/(eps*(||A||_oo*||x||_oo+||b||_oo)*N)= 0.0048373 ...... PASSED ================================================================================ T/V N NB P Q Time Gflops -------------------------------------------------------------------------------- WR00L4L2 16000 1060 2 2 26.05 1.048e+02 -------------------------------------------------------------------------------- ||Ax-b||_oo/(eps*(||A||_oo*||x||_oo+||b||_oo)*N)= 0.0048761 ...... PASSED ================================================================================ T/V N NB P Q Time Gflops -------------------------------------------------------------------------------- WR00L4L2 16000 1120 2 2 22.71 1.202e+02 -------------------------------------------------------------------------------- ||Ax-b||_oo/(eps*(||A||_oo*||x||_oo+||b||_oo)*N)= 0.0048042 ...... PASSED ================================================================================ T/V N NB P Q Time Gflops -------------------------------------------------------------------------------- WR00L4L2 16000 1130 2 2 26.43 1.033e+02 -------------------------------------------------------------------------------- ||Ax-b||_oo/(eps*(||A||_oo*||x||_oo+||b||_oo)*N)= 0.0048015 ...... PASSED ================================================================================ T/V N NB P Q Time Gflops -------------------------------------------------------------------------------- WR00L4L2 16000 1140 2 2 26.55 1.028e+02 -------------------------------------------------------------------------------- ||Ax-b||_oo/(eps*(||A||_oo*||x||_oo+||b||_oo)*N)= 0.0061187 ...... PASSED ================================================================================ Finished 9 tests with the following results: 9 tests completed and passed residual checks, 0 tests completed and failed residual checks, 0 tests skipped because of illegal input values. -------------------------------------------------------------------------------- End of Tests. ================================================================================ ======================== JOB MAP ======================== Data for node: Name: tesla2.localdomain Num procs: 4 Process OMPI jobid: [45524,1] Process rank: 0 Process OMPI jobid: [45524,1] Process rank: 1 Process OMPI jobid: [45524,1] Process rank: 2 Process OMPI jobid: [45524,1] Process rank: 3 ============================================================= ================================================================================ HPLinpack 2.0 -- High-Performance Linpack benchmark -- September 10, 2008 Written by A. Petitet and R. Clint Whaley, Innovative Computing Laboratory, UTK Modified by Piotr Luszczek, Innovative Computing Laboratory, UTK Modified by Julien Langou, University of Colorado Denver ================================================================================ An explanation of the input/output parameters follows: T/V : Wall time / encoded variant. N : The order of the coefficient matrix A. NB : The partitioning blocking factor. P : The number of process rows. Q : The number of process columns. Time : Time in seconds to solve the linear system. Gflops : Rate of execution for solving the linear system. The following parameter values will be used: N : 16000 NB : 1200 1280 1360 1440 PMAP : Row-major process mapping P : 2 Q : 2 PFACT : Left NBMIN : 2 NDIV : 4 RFACT : Left BCAST : 1ring DEPTH : 0 SWAP : Binary-exchange L1 : no-transposed form U : no-transposed form EQUIL : no ALIGN : 8 double precision words -------------------------------------------------------------------------------- - The matrix A is randomly generated for each test. - The following scaled residual check will be computed: ||Ax-b||_oo / ( eps * ( || x ||_oo * || A ||_oo + || b ||_oo ) * N ) - The relative machine precision (eps) is taken to be 1.110223e-16 - Computational tests pass if scaled residuals are less than 16.0 Assigning device 0 to process on node tesla2.localdomain rank 0 Assigning device 1 to process on node tesla2.localdomain rank 1 Assigning device 2 to process on node tesla2.localdomain rank 2 Assigning device 3 to process on node tesla2.localdomain rank 3 DTRSM split from environment variable 0.750000 DTRSM split from environment variable 0.750000 DGEMM split from environment variable 0.900000 DGEMM split from environment variable 0.900000 DTRSM split from environment variable 0.750000 DGEMM split from environment variable 0.900000 DTRSM split from environment variable 0.750000 DGEMM split from environment variable 0.900000 ================================================================================ T/V N NB P Q Time Gflops -------------------------------------------------------------------------------- WR00L4L2 16000 1200 2 2 22.84 1.196e+02 -------------------------------------------------------------------------------- ||Ax-b||_oo/(eps*(||A||_oo*||x||_oo+||b||_oo)*N)= 0.0049463 ...... PASSED ================================================================================ T/V N NB P Q Time Gflops -------------------------------------------------------------------------------- WR00L4L2 16000 1280 2 2 25.59 1.067e+02 -------------------------------------------------------------------------------- ||Ax-b||_oo/(eps*(||A||_oo*||x||_oo+||b||_oo)*N)= 0.0048939 ...... PASSED ================================================================================ T/V N NB P Q Time Gflops -------------------------------------------------------------------------------- WR00L4L2 16000 1360 2 2 23.55 1.160e+02 -------------------------------------------------------------------------------- ||Ax-b||_oo/(eps*(||A||_oo*||x||_oo+||b||_oo)*N)= 0.0049371 ...... PASSED ================================================================================ T/V N NB P Q Time Gflops -------------------------------------------------------------------------------- WR00L4L2 16000 1440 2 2 24.03 1.136e+02 -------------------------------------------------------------------------------- ||Ax-b||_oo/(eps*(||A||_oo*||x||_oo+||b||_oo)*N)= 0.0049782 ...... PASSED ================================================================================ Finished 4 tests with the following results: 4 tests completed and passed residual checks, 0 tests completed and failed residual checks, 0 tests skipped because of illegal input values. -------------------------------------------------------------------------------- End of Tests. ================================================================================ ======================== JOB MAP ======================== Data for node: Name: tesla2.localdomain Num procs: 4 Process OMPI jobid: [45547,1] Process rank: 0 Process OMPI jobid: [45547,1] Process rank: 1 Process OMPI jobid: [45547,1] Process rank: 2 Process OMPI jobid: [45547,1] Process rank: 3 ============================================================= ================================================================================ HPLinpack 2.0 -- High-Performance Linpack benchmark -- September 10, 2008 Written by A. Petitet and R. Clint Whaley, Innovative Computing Laboratory, UTK Modified by Piotr Luszczek, Innovative Computing Laboratory, UTK Modified by Julien Langou, University of Colorado Denver ================================================================================ An explanation of the input/output parameters follows: T/V : Wall time / encoded variant. N : The order of the coefficient matrix A. NB : The partitioning blocking factor. P : The number of process rows. Q : The number of process columns. Time : Time in seconds to solve the linear system. Gflops : Rate of execution for solving the linear system. The following parameter values will be used: N : 16000 NB : 1120 PMAP : Row-major process mapping P : 1 Q : 4 PFACT : Left NBMIN : 2 NDIV : 4 RFACT : Left BCAST : 1ring DEPTH : 0 SWAP : Binary-exchange L1 : no-transposed form U : no-transposed form EQUIL : no ALIGN : 8 double precision words -------------------------------------------------------------------------------- - The matrix A is randomly generated for each test. - The following scaled residual check will be computed: ||Ax-b||_oo / ( eps * ( || x ||_oo * || A ||_oo + || b ||_oo ) * N ) - The relative machine precision (eps) is taken to be 1.110223e-16 - Computational tests pass if scaled residuals are less than 16.0 Assigning device 0 to process on node tesla2.localdomain rank 0 Assigning device 3 to process on node tesla2.localdomain rank 3 Assigning device 1 to process on node tesla2.localdomain rank 1 Assigning device 2 to process on node tesla2.localdomain rank 2 DTRSM split from environment variable 0.750000 DGEMM split from environment variable 0.900000 DTRSM split from environment variable 0.750000 DTRSM split from environment variable 0.750000 DTRSM split from environment variable 0.750000 DGEMM split from environment variable 0.900000 DGEMM split from environment variable 0.900000 DGEMM split from environment variable 0.900000 ================================================================================ T/V N NB P Q Time Gflops -------------------------------------------------------------------------------- WR00L4L2 16000 1120 1 4 22.49 1.215e+02 -------------------------------------------------------------------------------- ||Ax-b||_oo/(eps*(||A||_oo*||x||_oo+||b||_oo)*N)= 0.0041107 ...... PASSED ================================================================================ Finished 1 tests with the following results: 1 tests completed and passed residual checks, 0 tests completed and failed residual checks, 0 tests skipped because of illegal input values. -------------------------------------------------------------------------------- End of Tests. ================================================================================ ======================== JOB MAP ======================== Data for node: Name: tesla2.localdomain Num procs: 4 Process OMPI jobid: [45453,1] Process rank: 0 Process OMPI jobid: [45453,1] Process rank: 1 Process OMPI jobid: [45453,1] Process rank: 2 Process OMPI jobid: [45453,1] Process rank: 3 ============================================================= ================================================================================ HPLinpack 2.0 -- High-Performance Linpack benchmark -- September 10, 2008 Written by A. Petitet and R. Clint Whaley, Innovative Computing Laboratory, UTK Modified by Piotr Luszczek, Innovative Computing Laboratory, UTK Modified by Julien Langou, University of Colorado Denver ================================================================================ An explanation of the input/output parameters follows: T/V : Wall time / encoded variant. N : The order of the coefficient matrix A. NB : The partitioning blocking factor. P : The number of process rows. Q : The number of process columns. Time : Time in seconds to solve the linear system. Gflops : Rate of execution for solving the linear system. The following parameter values will be used: N : 16000 NB : 1120 PMAP : Row-major process mapping P : 1 Q : 4 PFACT : Right NBMIN : 2 NDIV : 2 RFACT : Right BCAST : 1ringM DEPTH : 1 SWAP : Binary-exchange L1 : no-transposed form U : no-transposed form EQUIL : no ALIGN : 8 double precision words -------------------------------------------------------------------------------- - The matrix A is randomly generated for each test. - The following scaled residual check will be computed: ||Ax-b||_oo / ( eps * ( || x ||_oo * || A ||_oo + || b ||_oo ) * N ) - The relative machine precision (eps) is taken to be 1.110223e-16 - Computational tests pass if scaled residuals are less than 16.0 Assigning device 0 to process on node tesla2.localdomain rank 0 Assigning device 3 to process on node tesla2.localdomain rank 3 Assigning device 1 to process on node tesla2.localdomain rank 1 Assigning device 2 to process on node tesla2.localdomain rank 2 DTRSM split from environment variable 0.750000 DGEMM split from environment variable 0.900000 DTRSM split from environment variable 0.750000 DGEMM split from environment variable 0.900000 DTRSM split from environment variable 0.750000 DTRSM split from environment variable 0.750000 DGEMM split from environment variable 0.900000 DGEMM split from environment variable 0.900000 ================================================================================ T/V N NB P Q Time Gflops -------------------------------------------------------------------------------- WR11R2R2 16000 1120 1 4 18.44 1.481e+02 -------------------------------------------------------------------------------- ||Ax-b||_oo/(eps*(||A||_oo*||x||_oo+||b||_oo)*N)= 0.0039761 ...... PASSED ================================================================================ Finished 1 tests with the following results: 1 tests completed and passed residual checks, 0 tests completed and failed residual checks, 0 tests skipped because of illegal input values. -------------------------------------------------------------------------------- End of Tests. ================================================================================ ======================== JOB MAP ======================== Data for node: Name: tesla2.localdomain Num procs: 4 Process OMPI jobid: [45476,1] Process rank: 0 Process OMPI jobid: [45476,1] Process rank: 1 Process OMPI jobid: [45476,1] Process rank: 2 Process OMPI jobid: [45476,1] Process rank: 3 ============================================================= ================================================================================ HPLinpack 2.0 -- High-Performance Linpack benchmark -- September 10, 2008 Written by A. Petitet and R. Clint Whaley, Innovative Computing Laboratory, UTK Modified by Piotr Luszczek, Innovative Computing Laboratory, UTK Modified by Julien Langou, University of Colorado Denver ================================================================================ An explanation of the input/output parameters follows: T/V : Wall time / encoded variant. N : The order of the coefficient matrix A. NB : The partitioning blocking factor. P : The number of process rows. Q : The number of process columns. Time : Time in seconds to solve the linear system. Gflops : Rate of execution for solving the linear system. The following parameter values will be used: N : 56000 NB : 1120 PMAP : Row-major process mapping P : 1 Q : 4 PFACT : Right NBMIN : 2 NDIV : 2 RFACT : Right BCAST : 1ringM DEPTH : 1 SWAP : Binary-exchange L1 : no-transposed form U : no-transposed form EQUIL : no ALIGN : 8 double precision words -------------------------------------------------------------------------------- - The matrix A is randomly generated for each test. - The following scaled residual check will be computed: ||Ax-b||_oo / ( eps * ( || x ||_oo * || A ||_oo + || b ||_oo ) * N ) - The relative machine precision (eps) is taken to be 1.110223e-16 - Computational tests pass if scaled residuals are less than 16.0 Assigning device 0 to process on node tesla2.localdomain rank 0 Assigning device 2 to process on node tesla2.localdomain rank 2 Assigning device 1 to process on node tesla2.localdomain rank 1 Assigning device 3 to process on node tesla2.localdomain rank 3 DTRSM split from environment variable 0.750000 DGEMM split from environment variable 0.900000 DTRSM split from environment variable 0.750000 DGEMM split from environment variable 0.900000 DTRSM split from environment variable 0.750000 DTRSM split from environment variable 0.750000 DGEMM split from environment variable 0.900000 DGEMM split from environment variable 0.900000 ================================================================================ T/V N NB P Q Time Gflops -------------------------------------------------------------------------------- WR11R2R2 56000 1120 1 4 749.26 1.563e+02 -------------------------------------------------------------------------------- ||Ax-b||_oo/(eps*(||A||_oo*||x||_oo+||b||_oo)*N)= 327689498.1566594 ...... FAILED ||Ax-b||_oo . . . . . . . . . . . . . . . . . = 270.717850 ||A||_oo . . . . . . . . . . . . . . . . . . . = 14147.398512 ||A||_1 . . . . . . . . . . . . . . . . . . . = 14144.374333 ||x||_oo . . . . . . . . . . . . . . . . . . . = 9.392428 ||x||_1 . . . . . . . . . . . . . . . . . . . = 26204.111995 ||b||_oo . . . . . . . . . . . . . . . . . . . = 0.500000 ================================================================================ Finished 1 tests with the following results: 0 tests completed and passed residual checks, 1 tests completed and failed residual checks, 0 tests skipped because of illegal input values. -------------------------------------------------------------------------------- End of Tests. ================================================================================ ======================== JOB MAP ======================== Data for node: Name: tesla2.localdomain Num procs: 4 Process OMPI jobid: [45390,1] Process rank: 0 Process OMPI jobid: [45390,1] Process rank: 1 Process OMPI jobid: [45390,1] Process rank: 2 Process OMPI jobid: [45390,1] Process rank: 3 ============================================================= ================================================================================ HPLinpack 2.0 -- High-Performance Linpack benchmark -- September 10, 2008 Written by A. Petitet and R. Clint Whaley, Innovative Computing Laboratory, UTK Modified by Piotr Luszczek, Innovative Computing Laboratory, UTK Modified by Julien Langou, University of Colorado Denver ================================================================================ An explanation of the input/output parameters follows: T/V : Wall time / encoded variant. N : The order of the coefficient matrix A. NB : The partitioning blocking factor. P : The number of process rows. Q : The number of process columns. Time : Time in seconds to solve the linear system. Gflops : Rate of execution for solving the linear system. The following parameter values will be used: N : 56000 NB : 1120 PMAP : Row-major process mapping P : 2 Q : 2 PFACT : Right NBMIN : 2 NDIV : 2 RFACT : Right BCAST : 1ringM DEPTH : 1 SWAP : Binary-exchange L1 : no-transposed form U : no-transposed form EQUIL : no ALIGN : 8 double precision words -------------------------------------------------------------------------------- - The matrix A is randomly generated for each test. - The following scaled residual check will be computed: ||Ax-b||_oo / ( eps * ( || x ||_oo * || A ||_oo + || b ||_oo ) * N ) - The relative machine precision (eps) is taken to be 1.110223e-16 - Computational tests pass if scaled residuals are less than 16.0 Assigning device 0 to process on node tesla2.localdomain rank 0 Assigning device 1 to process on node tesla2.localdomain rank 1 Assigning device 3 to process on node tesla2.localdomain rank 3 Assigning device 2 to process on node tesla2.localdomain rank 2 DTRSM split from environment variable 0.750000 DGEMM split from environment variable 0.900000 DTRSM split from environment variable 0.750000 DGEMM split from environment variable 0.900000 DTRSM split from environment variable 0.750000 DGEMM split from environment variable 0.900000 DTRSM split from environment variable 0.750000 DGEMM split from environment variable 0.900000 ================================================================================ T/V N NB P Q Time Gflops -------------------------------------------------------------------------------- WR11R2R2 56000 1120 2 2 517.18 2.264e+02 -------------------------------------------------------------------------------- ||Ax-b||_oo/(eps*(||A||_oo*||x||_oo+||b||_oo)*N)= 0.0038950 ...... PASSED ================================================================================ Finished 1 tests with the following results: 1 tests completed and passed residual checks, 0 tests completed and failed residual checks, 0 tests skipped because of illegal input values. -------------------------------------------------------------------------------- End of Tests. ================================================================================ ======================== JOB MAP ======================== Data for node: Name: tesla2.localdomain Num procs: 4 Process OMPI jobid: [45432,1] Process rank: 0 Process OMPI jobid: [45432,1] Process rank: 1 Process OMPI jobid: [45432,1] Process rank: 2 Process OMPI jobid: [45432,1] Process rank: 3 ============================================================= ================================================================================ HPLinpack 2.0 -- High-Performance Linpack benchmark -- September 10, 2008 Written by A. Petitet and R. Clint Whaley, Innovative Computing Laboratory, UTK Modified by Piotr Luszczek, Innovative Computing Laboratory, UTK Modified by Julien Langou, University of Colorado Denver ================================================================================ An explanation of the input/output parameters follows: T/V : Wall time / encoded variant. N : The order of the coefficient matrix A. NB : The partitioning blocking factor. P : The number of process rows. Q : The number of process columns. Time : Time in seconds to solve the linear system. Gflops : Rate of execution for solving the linear system. The following parameter values will be used: N : 60000 NB : 1120 PMAP : Row-major process mapping P : 2 Q : 2 PFACT : Right NBMIN : 2 NDIV : 2 RFACT : Right BCAST : 1ringM DEPTH : 1 SWAP : Binary-exchange L1 : no-transposed form U : no-transposed form EQUIL : no ALIGN : 8 double precision words -------------------------------------------------------------------------------- - The matrix A is randomly generated for each test. - The following scaled residual check will be computed: ||Ax-b||_oo / ( eps * ( || x ||_oo * || A ||_oo + || b ||_oo ) * N ) - The relative machine precision (eps) is taken to be 1.110223e-16 - Computational tests pass if scaled residuals are less than 16.0 Assigning device 0 to process on node tesla2.localdomain rank 0 Assigning device 1 to process on node tesla2.localdomain rank 1 Assigning device 2 to process on node tesla2.localdomain rank 2 Assigning device 3 to process on node tesla2.localdomain rank 3 DTRSM split from environment variable 0.750000 DTRSM split from environment variable 0.750000 DGEMM split from environment variable 0.900000 DGEMM split from environment variable 0.900000 DTRSM split from environment variable 0.750000 DGEMM split from environment variable 0.900000 DTRSM split from environment variable 0.750000 DGEMM split from environment variable 0.900000 ================================================================================ T/V N NB P Q Time Gflops -------------------------------------------------------------------------------- WR11R2R2 60000 1120 2 2 624.56 2.306e+02 -------------------------------------------------------------------------------- ||Ax-b||_oo/(eps*(||A||_oo*||x||_oo+||b||_oo)*N)= 0.0044876 ...... PASSED ================================================================================ Finished 1 tests with the following results: 1 tests completed and passed residual checks, 0 tests completed and failed residual checks, 0 tests skipped because of illegal input values. -------------------------------------------------------------------------------- End of Tests. ================================================================================ ======================== JOB MAP ======================== Data for node: Name: tesla2.localdomain Num procs: 4 Process OMPI jobid: [45349,1] Process rank: 0 Process OMPI jobid: [45349,1] Process rank: 1 Process OMPI jobid: [45349,1] Process rank: 2 Process OMPI jobid: [45349,1] Process rank: 3 ============================================================= ================================================================================ HPLinpack 2.0 -- High-Performance Linpack benchmark -- September 10, 2008 Written by A. Petitet and R. Clint Whaley, Innovative Computing Laboratory, UTK Modified by Piotr Luszczek, Innovative Computing Laboratory, UTK Modified by Julien Langou, University of Colorado Denver ================================================================================ An explanation of the input/output parameters follows: T/V : Wall time / encoded variant. N : The order of the coefficient matrix A. NB : The partitioning blocking factor. P : The number of process rows. Q : The number of process columns. Time : Time in seconds to solve the linear system. Gflops : Rate of execution for solving the linear system. The following parameter values will be used: N : 16000 NB : 1120 PMAP : Row-major process mapping P : 2 Q : 2 PFACT : Right NBMIN : 2 NDIV : 2 RFACT : Right BCAST : 1ringM DEPTH : 1 SWAP : Binary-exchange L1 : no-transposed form U : no-transposed form EQUIL : no ALIGN : 8 double precision words -------------------------------------------------------------------------------- - The matrix A is randomly generated for each test. - The following scaled residual check will be computed: ||Ax-b||_oo / ( eps * ( || x ||_oo * || A ||_oo + || b ||_oo ) * N ) - The relative machine precision (eps) is taken to be 1.110223e-16 - Computational tests pass if scaled residuals are less than 16.0 Assigning device 0 to process on node tesla2.localdomain rank 0 Assigning device 2 to process on node tesla2.localdomain rank 2 Assigning device 3 to process on node tesla2.localdomain rank 3 Assigning device 1 to process on node tesla2.localdomain rank 1 DTRSM split from environment variable 0.750000 DGEMM split from environment variable 0.900000 DTRSM split from environment variable 0.750000 DGEMM split from environment variable 0.900000 DTRSM split from environment variable 0.750000 DGEMM split from environment variable 0.900000 DTRSM split from environment variable 0.750000 DGEMM split from environment variable 0.900000 ================================================================================ T/V N NB P Q Time Gflops -------------------------------------------------------------------------------- WR11R2R2 16000 1120 2 2 21.25 1.285e+02 -------------------------------------------------------------------------------- ||Ax-b||_oo/(eps*(||A||_oo*||x||_oo+||b||_oo)*N)= 0.0057905 ...... PASSED ================================================================================ Finished 1 tests with the following results: 1 tests completed and passed residual checks, 0 tests completed and failed residual checks, 0 tests skipped because of illegal input values. -------------------------------------------------------------------------------- End of Tests. ================================================================================ ======================== JOB MAP ======================== Data for node: Name: tesla2.localdomain Num procs: 4 Process OMPI jobid: [45371,1] Process rank: 0 Process OMPI jobid: [45371,1] Process rank: 1 Process OMPI jobid: [45371,1] Process rank: 2 Process OMPI jobid: [45371,1] Process rank: 3 ============================================================= ================================================================================ HPLinpack 2.0 -- High-Performance Linpack benchmark -- September 10, 2008 Written by A. Petitet and R. Clint Whaley, Innovative Computing Laboratory, UTK Modified by Piotr Luszczek, Innovative Computing Laboratory, UTK Modified by Julien Langou, University of Colorado Denver ================================================================================ An explanation of the input/output parameters follows: T/V : Wall time / encoded variant. N : The order of the coefficient matrix A. NB : The partitioning blocking factor. P : The number of process rows. Q : The number of process columns. Time : Time in seconds to solve the linear system. Gflops : Rate of execution for solving the linear system. The following parameter values will be used: N : 16000 NB : 1120 PMAP : Row-major process mapping P : 2 Q : 2 PFACT : Right NBMIN : 2 NDIV : 2 RFACT : Right BCAST : 1ringM DEPTH : 1 SWAP : Binary-exchange L1 : no-transposed form U : no-transposed form EQUIL : no ALIGN : 8 double precision words -------------------------------------------------------------------------------- - The matrix A is randomly generated for each test. - The following scaled residual check will be computed: ||Ax-b||_oo / ( eps * ( || x ||_oo * || A ||_oo + || b ||_oo ) * N ) - The relative machine precision (eps) is taken to be 1.110223e-16 - Computational tests pass if scaled residuals are less than 16.0 Assigning device 0 to process on node tesla2.localdomain rank 0 Assigning device 2 to process on node tesla2.localdomain rank 2 Assigning device 3 to process on node tesla2.localdomain rank 3 Assigning device 1 to process on node tesla2.localdomain rank 1 DTRSM split from environment variable 0.550000 DGEMM split from environment variable 0.900000 DTRSM split from environment variable 0.550000 DGEMM split from environment variable 0.900000 DTRSM split from environment variable 0.550000 DGEMM split from environment variable 0.900000 DTRSM split from environment variable 0.550000 DGEMM split from environment variable 0.900000 ================================================================================ T/V N NB P Q Time Gflops -------------------------------------------------------------------------------- WR11R2R2 16000 1120 2 2 22.37 1.221e+02 -------------------------------------------------------------------------------- ||Ax-b||_oo/(eps*(||A||_oo*||x||_oo+||b||_oo)*N)= 0.0042796 ...... PASSED ================================================================================ Finished 1 tests with the following results: 1 tests completed and passed residual checks, 0 tests completed and failed residual checks, 0 tests skipped because of illegal input values. -------------------------------------------------------------------------------- End of Tests. ================================================================================ ======================== JOB MAP ======================== Data for node: Name: tesla2.localdomain Num procs: 4 Process OMPI jobid: [45264,1] Process rank: 0 Process OMPI jobid: [45264,1] Process rank: 1 Process OMPI jobid: [45264,1] Process rank: 2 Process OMPI jobid: [45264,1] Process rank: 3 ============================================================= ================================================================================ HPLinpack 2.0 -- High-Performance Linpack benchmark -- September 10, 2008 Written by A. Petitet and R. Clint Whaley, Innovative Computing Laboratory, UTK Modified by Piotr Luszczek, Innovative Computing Laboratory, UTK Modified by Julien Langou, University of Colorado Denver ================================================================================ An explanation of the input/output parameters follows: T/V : Wall time / encoded variant. N : The order of the coefficient matrix A. NB : The partitioning blocking factor. P : The number of process rows. Q : The number of process columns. Time : Time in seconds to solve the linear system. Gflops : Rate of execution for solving the linear system. The following parameter values will be used: N : 16000 NB : 1120 PMAP : Row-major process mapping P : 2 Q : 2 PFACT : Right NBMIN : 2 NDIV : 2 RFACT : Right BCAST : 1ringM DEPTH : 1 SWAP : Binary-exchange L1 : no-transposed form U : no-transposed form EQUIL : no ALIGN : 8 double precision words -------------------------------------------------------------------------------- - The matrix A is randomly generated for each test. - The following scaled residual check will be computed: ||Ax-b||_oo / ( eps * ( || x ||_oo * || A ||_oo + || b ||_oo ) * N ) - The relative machine precision (eps) is taken to be 1.110223e-16 - Computational tests pass if scaled residuals are less than 16.0 Assigning device 0 to process on node tesla2.localdomain rank 0 Assigning device 3 to process on node tesla2.localdomain rank 3 Assigning device 1 to process on node tesla2.localdomain rank 1 Assigning device 2 to process on node tesla2.localdomain rank 2 DTRSM split from environment variable 0.350000 DTRSM split from environment variable 0.350000 DGEMM split from environment variable 0.900000 DGEMM split from environment variable 0.900000 DTRSM split from environment variable 0.350000 DTRSM split from environment variable 0.350000 DGEMM split from environment variable 0.900000 DGEMM split from environment variable 0.900000 ================================================================================ T/V N NB P Q Time Gflops -------------------------------------------------------------------------------- WR11R2R2 16000 1120 2 2 23.09 1.183e+02 -------------------------------------------------------------------------------- ||Ax-b||_oo/(eps*(||A||_oo*||x||_oo+||b||_oo)*N)= 0.0056988 ...... PASSED ================================================================================ Finished 1 tests with the following results: 1 tests completed and passed residual checks, 0 tests completed and failed residual checks, 0 tests skipped because of illegal input values. -------------------------------------------------------------------------------- End of Tests. ================================================================================ ======================== JOB MAP ======================== Data for node: Name: tesla2.localdomain Num procs: 4 Process OMPI jobid: [45286,1] Process rank: 0 Process OMPI jobid: [45286,1] Process rank: 1 Process OMPI jobid: [45286,1] Process rank: 2 Process OMPI jobid: [45286,1] Process rank: 3 ============================================================= ================================================================================ HPLinpack 2.0 -- High-Performance Linpack benchmark -- September 10, 2008 Written by A. Petitet and R. Clint Whaley, Innovative Computing Laboratory, UTK Modified by Piotr Luszczek, Innovative Computing Laboratory, UTK Modified by Julien Langou, University of Colorado Denver ================================================================================ An explanation of the input/output parameters follows: T/V : Wall time / encoded variant. N : The order of the coefficient matrix A. NB : The partitioning blocking factor. P : The number of process rows. Q : The number of process columns. Time : Time in seconds to solve the linear system. Gflops : Rate of execution for solving the linear system. The following parameter values will be used: N : 16000 NB : 1120 PMAP : Row-major process mapping P : 2 Q : 2 PFACT : Right NBMIN : 2 NDIV : 2 RFACT : Right BCAST : 1ringM DEPTH : 1 SWAP : Binary-exchange L1 : no-transposed form U : no-transposed form EQUIL : no ALIGN : 8 double precision words -------------------------------------------------------------------------------- - The matrix A is randomly generated for each test. - The following scaled residual check will be computed: ||Ax-b||_oo / ( eps * ( || x ||_oo * || A ||_oo + || b ||_oo ) * N ) - The relative machine precision (eps) is taken to be 1.110223e-16 - Computational tests pass if scaled residuals are less than 16.0 Assigning device 3 to process on node tesla2.localdomain rank 3 Assigning device 0 to process on node tesla2.localdomain rank 0 Assigning device 1 to process on node tesla2.localdomain rank 1 Assigning device 2 to process on node tesla2.localdomain rank 2 DTRSM split from environment variable 0.850000 DTRSM split from environment variable 0.850000 DGEMM split from environment variable 0.900000 DGEMM split from environment variable 0.900000 DTRSM split from environment variable 0.850000 DGEMM split from environment variable 0.900000 DTRSM split from environment variable 0.850000 DGEMM split from environment variable 0.900000 ================================================================================ T/V N NB P Q Time Gflops -------------------------------------------------------------------------------- WR11R2R2 16000 1120 2 2 21.49 1.271e+02 -------------------------------------------------------------------------------- ||Ax-b||_oo/(eps*(||A||_oo*||x||_oo+||b||_oo)*N)= 0.0046317 ...... PASSED ================================================================================ Finished 1 tests with the following results: 1 tests completed and passed residual checks, 0 tests completed and failed residual checks, 0 tests skipped because of illegal input values. -------------------------------------------------------------------------------- End of Tests. ================================================================================ ======================== JOB MAP ======================== Data for node: Name: tesla2.localdomain Num procs: 4 Process OMPI jobid: [45190,1] Process rank: 0 Process OMPI jobid: [45190,1] Process rank: 1 Process OMPI jobid: [45190,1] Process rank: 2 Process OMPI jobid: [45190,1] Process rank: 3 ============================================================= ================================================================================ HPLinpack 2.0 -- High-Performance Linpack benchmark -- September 10, 2008 Written by A. Petitet and R. Clint Whaley, Innovative Computing Laboratory, UTK Modified by Piotr Luszczek, Innovative Computing Laboratory, UTK Modified by Julien Langou, University of Colorado Denver ================================================================================ An explanation of the input/output parameters follows: T/V : Wall time / encoded variant. N : The order of the coefficient matrix A. NB : The partitioning blocking factor. P : The number of process rows. Q : The number of process columns. Time : Time in seconds to solve the linear system. Gflops : Rate of execution for solving the linear system. The following parameter values will be used: N : 16000 NB : 1120 PMAP : Row-major process mapping P : 2 Q : 2 PFACT : Right NBMIN : 2 NDIV : 2 RFACT : Right BCAST : 1ringM DEPTH : 1 SWAP : Binary-exchange L1 : no-transposed form U : no-transposed form EQUIL : no ALIGN : 8 double precision words -------------------------------------------------------------------------------- - The matrix A is randomly generated for each test. - The following scaled residual check will be computed: ||Ax-b||_oo / ( eps * ( || x ||_oo * || A ||_oo + || b ||_oo ) * N ) - The relative machine precision (eps) is taken to be 1.110223e-16 - Computational tests pass if scaled residuals are less than 16.0 Assigning device 0 to process on node tesla2.localdomain rank 0 Assigning device 1 to process on node tesla2.localdomain rank 1 Assigning device 2 to process on node tesla2.localdomain rank 2 Assigning device 3 to process on node tesla2.localdomain rank 3 DTRSM split from environment variable 0.750000 DTRSM split from environment variable 0.750000 DGEMM split from environment variable 0.900000 DGEMM split from environment variable 0.900000 DTRSM split from environment variable 0.750000 DGEMM split from environment variable 0.900000 DTRSM split from environment variable 0.750000 DGEMM split from environment variable 0.900000 ================================================================================ T/V N NB P Q Time Gflops -------------------------------------------------------------------------------- WR11R2R2 16000 1120 2 2 21.34 1.280e+02 -------------------------------------------------------------------------------- ||Ax-b||_oo/(eps*(||A||_oo*||x||_oo+||b||_oo)*N)= 0.0048646 ...... PASSED ================================================================================ Finished 1 tests with the following results: 1 tests completed and passed residual checks, 0 tests completed and failed residual checks, 0 tests skipped because of illegal input values. -------------------------------------------------------------------------------- End of Tests. ================================================================================ ======================== JOB MAP ======================== Data for node: Name: tesla2.localdomain Num procs: 4 Process OMPI jobid: [45212,1] Process rank: 0 Process OMPI jobid: [45212,1] Process rank: 1 Process OMPI jobid: [45212,1] Process rank: 2 Process OMPI jobid: [45212,1] Process rank: 3 ============================================================= ================================================================================ HPLinpack 2.0 -- High-Performance Linpack benchmark -- September 10, 2008 Written by A. Petitet and R. Clint Whaley, Innovative Computing Laboratory, UTK Modified by Piotr Luszczek, Innovative Computing Laboratory, UTK Modified by Julien Langou, University of Colorado Denver ================================================================================ An explanation of the input/output parameters follows: T/V : Wall time / encoded variant. N : The order of the coefficient matrix A. NB : The partitioning blocking factor. P : The number of process rows. Q : The number of process columns. Time : Time in seconds to solve the linear system. Gflops : Rate of execution for solving the linear system. The following parameter values will be used: N : 16000 NB : 1120 PMAP : Row-major process mapping P : 2 Q : 2 PFACT : Right NBMIN : 2 NDIV : 2 RFACT : Right BCAST : 1ringM DEPTH : 1 SWAP : Binary-exchange L1 : no-transposed form U : no-transposed form EQUIL : no ALIGN : 8 double precision words -------------------------------------------------------------------------------- - The matrix A is randomly generated for each test. - The following scaled residual check will be computed: ||Ax-b||_oo / ( eps * ( || x ||_oo * || A ||_oo + || b ||_oo ) * N ) - The relative machine precision (eps) is taken to be 1.110223e-16 - Computational tests pass if scaled residuals are less than 16.0 Assigning device 0 to process on node tesla2.localdomain rank 0 Assigning device 1 to process on node tesla2.localdomain rank 1 Assigning device 2 to process on node tesla2.localdomain rank 2 Assigning device 3 to process on node tesla2.localdomain rank 3 DTRSM split from environment variable 0.650000 DTRSM split from environment variable 0.650000 DGEMM split from environment variable 0.900000 DGEMM split from environment variable 0.900000 DTRSM split from environment variable 0.650000 DTRSM split from environment variable 0.650000 DGEMM split from environment variable 0.900000 DGEMM split from environment variable 0.900000 ================================================================================ T/V N NB P Q Time Gflops -------------------------------------------------------------------------------- WR11R2R2 16000 1120 2 2 21.77 1.255e+02 -------------------------------------------------------------------------------- ||Ax-b||_oo/(eps*(||A||_oo*||x||_oo+||b||_oo)*N)= 0.0048376 ...... PASSED ================================================================================ Finished 1 tests with the following results: 1 tests completed and passed residual checks, 0 tests completed and failed residual checks, 0 tests skipped because of illegal input values. -------------------------------------------------------------------------------- End of Tests. ================================================================================ ======================== JOB MAP ======================== Data for node: Name: tesla2.localdomain Num procs: 4 Process OMPI jobid: [45234,1] Process rank: 0 Process OMPI jobid: [45234,1] Process rank: 1 Process OMPI jobid: [45234,1] Process rank: 2 Process OMPI jobid: [45234,1] Process rank: 3 ============================================================= ================================================================================ HPLinpack 2.0 -- High-Performance Linpack benchmark -- September 10, 2008 Written by A. Petitet and R. Clint Whaley, Innovative Computing Laboratory, UTK Modified by Piotr Luszczek, Innovative Computing Laboratory, UTK Modified by Julien Langou, University of Colorado Denver ================================================================================ An explanation of the input/output parameters follows: T/V : Wall time / encoded variant. N : The order of the coefficient matrix A. NB : The partitioning blocking factor. P : The number of process rows. Q : The number of process columns. Time : Time in seconds to solve the linear system. Gflops : Rate of execution for solving the linear system. The following parameter values will be used: N : 16000 NB : 1120 PMAP : Row-major process mapping P : 2 Q : 2 PFACT : Right NBMIN : 2 NDIV : 2 RFACT : Right BCAST : 1ringM DEPTH : 1 SWAP : Binary-exchange L1 : no-transposed form U : no-transposed form EQUIL : no ALIGN : 8 double precision words -------------------------------------------------------------------------------- - The matrix A is randomly generated for each test. - The following scaled residual check will be computed: ||Ax-b||_oo / ( eps * ( || x ||_oo * || A ||_oo + || b ||_oo ) * N ) - The relative machine precision (eps) is taken to be 1.110223e-16 - Computational tests pass if scaled residuals are less than 16.0 Assigning device 0 to process on node tesla2.localdomain rank 0 Assigning device 1 to process on node tesla2.localdomain rank 1 Assigning device 2 to process on node tesla2.localdomain rank 2 Assigning device 3 to process on node tesla2.localdomain rank 3 DTRSM split from environment variable 0.700000 DTRSM split from environment variable 0.700000 DGEMM split from environment variable 0.900000 DGEMM split from environment variable 0.900000 DTRSM split from environment variable 0.700000 DGEMM split from environment variable 0.900000 DTRSM split from environment variable 0.700000 DGEMM split from environment variable 0.900000 ================================================================================ T/V N NB P Q Time Gflops -------------------------------------------------------------------------------- WR11R2R2 16000 1120 2 2 21.72 1.258e+02 -------------------------------------------------------------------------------- ||Ax-b||_oo/(eps*(||A||_oo*||x||_oo+||b||_oo)*N)= 0.0049889 ...... PASSED ================================================================================ Finished 1 tests with the following results: 1 tests completed and passed residual checks, 0 tests completed and failed residual checks, 0 tests skipped because of illegal input values. -------------------------------------------------------------------------------- End of Tests. ================================================================================ ======================== JOB MAP ======================== Data for node: Name: tesla2.localdomain Num procs: 4 Process OMPI jobid: [45139,1] Process rank: 0 Process OMPI jobid: [45139,1] Process rank: 1 Process OMPI jobid: [45139,1] Process rank: 2 Process OMPI jobid: [45139,1] Process rank: 3 ============================================================= ================================================================================ HPLinpack 2.0 -- High-Performance Linpack benchmark -- September 10, 2008 Written by A. Petitet and R. Clint Whaley, Innovative Computing Laboratory, UTK Modified by Piotr Luszczek, Innovative Computing Laboratory, UTK Modified by Julien Langou, University of Colorado Denver ================================================================================ An explanation of the input/output parameters follows: T/V : Wall time / encoded variant. N : The order of the coefficient matrix A. NB : The partitioning blocking factor. P : The number of process rows. Q : The number of process columns. Time : Time in seconds to solve the linear system. Gflops : Rate of execution for solving the linear system. The following parameter values will be used: N : 16000 NB : 1120 PMAP : Row-major process mapping P : 2 Q : 2 PFACT : Right NBMIN : 2 NDIV : 2 RFACT : Right BCAST : 1ringM DEPTH : 1 SWAP : Binary-exchange L1 : no-transposed form U : no-transposed form EQUIL : no ALIGN : 8 double precision words -------------------------------------------------------------------------------- - The matrix A is randomly generated for each test. - The following scaled residual check will be computed: ||Ax-b||_oo / ( eps * ( || x ||_oo * || A ||_oo + || b ||_oo ) * N ) - The relative machine precision (eps) is taken to be 1.110223e-16 - Computational tests pass if scaled residuals are less than 16.0 Assigning device 0 to process on node tesla2.localdomain rank 0 Assigning device 3 to process on node tesla2.localdomain rank 3 Assigning device 2 to process on node tesla2.localdomain rank 2 Assigning device 1 to process on node tesla2.localdomain rank 1 DTRSM split from environment variable 0.760000 DGEMM split from environment variable 0.900000 DTRSM split from environment variable 0.760000 DGEMM split from environment variable 0.900000 DTRSM split from environment variable 0.760000 DGEMM split from environment variable 0.900000 DTRSM split from environment variable 0.760000 DGEMM split from environment variable 0.900000 ================================================================================ T/V N NB P Q Time Gflops -------------------------------------------------------------------------------- WR11R2R2 16000 1120 2 2 21.32 1.281e+02 -------------------------------------------------------------------------------- ||Ax-b||_oo/(eps*(||A||_oo*||x||_oo+||b||_oo)*N)= 0.0051775 ...... PASSED ================================================================================ Finished 1 tests with the following results: 1 tests completed and passed residual checks, 0 tests completed and failed residual checks, 0 tests skipped because of illegal input values. -------------------------------------------------------------------------------- End of Tests. ================================================================================ ======================== JOB MAP ======================== Data for node: Name: tesla2.localdomain Num procs: 4 Process OMPI jobid: [45161,1] Process rank: 0 Process OMPI jobid: [45161,1] Process rank: 1 Process OMPI jobid: [45161,1] Process rank: 2 Process OMPI jobid: [45161,1] Process rank: 3 ============================================================= ================================================================================ HPLinpack 2.0 -- High-Performance Linpack benchmark -- September 10, 2008 Written by A. Petitet and R. Clint Whaley, Innovative Computing Laboratory, UTK Modified by Piotr Luszczek, Innovative Computing Laboratory, UTK Modified by Julien Langou, University of Colorado Denver ================================================================================ An explanation of the input/output parameters follows: T/V : Wall time / encoded variant. N : The order of the coefficient matrix A. NB : The partitioning blocking factor. P : The number of process rows. Q : The number of process columns. Time : Time in seconds to solve the linear system. Gflops : Rate of execution for solving the linear system. The following parameter values will be used: N : 16000 NB : 1120 PMAP : Row-major process mapping P : 2 Q : 2 PFACT : Right NBMIN : 2 NDIV : 2 RFACT : Right BCAST : 1ringM DEPTH : 1 SWAP : Binary-exchange L1 : no-transposed form U : no-transposed form EQUIL : no ALIGN : 8 double precision words -------------------------------------------------------------------------------- - The matrix A is randomly generated for each test. - The following scaled residual check will be computed: ||Ax-b||_oo / ( eps * ( || x ||_oo * || A ||_oo + || b ||_oo ) * N ) - The relative machine precision (eps) is taken to be 1.110223e-16 - Computational tests pass if scaled residuals are less than 16.0 Assigning device 0 to process on node tesla2.localdomain rank 0 Assigning device 2 to process on node tesla2.localdomain rank 2 Assigning device 3 to process on node tesla2.localdomain rank 3 Assigning device 1 to process on node tesla2.localdomain rank 1 DTRSM split from environment variable 0.780000 DTRSM split from environment variable 0.780000 DGEMM split from environment variable 0.900000 DGEMM split from environment variable 0.900000 DTRSM split from environment variable 0.780000 DGEMM split from environment variable 0.900000 DTRSM split from environment variable 0.780000 DGEMM split from environment variable 0.900000 ================================================================================ T/V N NB P Q Time Gflops -------------------------------------------------------------------------------- WR11R2R2 16000 1120 2 2 21.35 1.279e+02 -------------------------------------------------------------------------------- ||Ax-b||_oo/(eps*(||A||_oo*||x||_oo+||b||_oo)*N)= 0.0045566 ...... PASSED ================================================================================ Finished 1 tests with the following results: 1 tests completed and passed residual checks, 0 tests completed and failed residual checks, 0 tests skipped because of illegal input values. -------------------------------------------------------------------------------- End of Tests. ================================================================================ ======================== JOB MAP ======================== Data for node: Name: tesla2.localdomain Num procs: 4 Process OMPI jobid: [45182,1] Process rank: 0 Process OMPI jobid: [45182,1] Process rank: 1 Process OMPI jobid: [45182,1] Process rank: 2 Process OMPI jobid: [45182,1] Process rank: 3 ============================================================= ================================================================================ HPLinpack 2.0 -- High-Performance Linpack benchmark -- September 10, 2008 Written by A. Petitet and R. Clint Whaley, Innovative Computing Laboratory, UTK Modified by Piotr Luszczek, Innovative Computing Laboratory, UTK Modified by Julien Langou, University of Colorado Denver ================================================================================ An explanation of the input/output parameters follows: T/V : Wall time / encoded variant. N : The order of the coefficient matrix A. NB : The partitioning blocking factor. P : The number of process rows. Q : The number of process columns. Time : Time in seconds to solve the linear system. Gflops : Rate of execution for solving the linear system. The following parameter values will be used: N : 16000 NB : 1120 PMAP : Row-major process mapping P : 2 Q : 2 PFACT : Right NBMIN : 2 NDIV : 2 RFACT : Right BCAST : 1ringM DEPTH : 1 SWAP : Binary-exchange L1 : no-transposed form U : no-transposed form EQUIL : no ALIGN : 8 double precision words -------------------------------------------------------------------------------- - The matrix A is randomly generated for each test. - The following scaled residual check will be computed: ||Ax-b||_oo / ( eps * ( || x ||_oo * || A ||_oo + || b ||_oo ) * N ) - The relative machine precision (eps) is taken to be 1.110223e-16 - Computational tests pass if scaled residuals are less than 16.0 Assigning device 0 to process on node tesla2.localdomain rank 0 Assigning device 1 to process on node tesla2.localdomain rank 1 Assigning device 2 to process on node tesla2.localdomain rank 2 Assigning device 3 to process on node tesla2.localdomain rank 3 DTRSM split from environment variable 0.740000 DGEMM split from environment variable 0.900000 DTRSM split from environment variable 0.740000 DGEMM split from environment variable 0.900000 DTRSM split from environment variable 0.740000 DGEMM split from environment variable 0.900000 DTRSM split from environment variable 0.740000 DGEMM split from environment variable 0.900000 ================================================================================ T/V N NB P Q Time Gflops -------------------------------------------------------------------------------- WR11R2R2 16000 1120 2 2 21.33 1.280e+02 -------------------------------------------------------------------------------- ||Ax-b||_oo/(eps*(||A||_oo*||x||_oo+||b||_oo)*N)= 0.0048610 ...... PASSED ================================================================================ Finished 1 tests with the following results: 1 tests completed and passed residual checks, 0 tests completed and failed residual checks, 0 tests skipped because of illegal input values. -------------------------------------------------------------------------------- End of Tests. ================================================================================ ======================== JOB MAP ======================== Data for node: Name: tesla2.localdomain Num procs: 4 Process OMPI jobid: [45086,1] Process rank: 0 Process OMPI jobid: [45086,1] Process rank: 1 Process OMPI jobid: [45086,1] Process rank: 2 Process OMPI jobid: [45086,1] Process rank: 3 ============================================================= ================================================================================ HPLinpack 2.0 -- High-Performance Linpack benchmark -- September 10, 2008 Written by A. Petitet and R. Clint Whaley, Innovative Computing Laboratory, UTK Modified by Piotr Luszczek, Innovative Computing Laboratory, UTK Modified by Julien Langou, University of Colorado Denver ================================================================================ An explanation of the input/output parameters follows: T/V : Wall time / encoded variant. N : The order of the coefficient matrix A. NB : The partitioning blocking factor. P : The number of process rows. Q : The number of process columns. Time : Time in seconds to solve the linear system. Gflops : Rate of execution for solving the linear system. The following parameter values will be used: N : 16000 NB : 1120 PMAP : Row-major process mapping P : 2 Q : 2 PFACT : Left NBMIN : 2 NDIV : 2 RFACT : Left BCAST : 1ring DEPTH : 0 SWAP : Binary-exchange L1 : no-transposed form U : no-transposed form EQUIL : no ALIGN : 8 double precision words -------------------------------------------------------------------------------- - The matrix A is randomly generated for each test. - The following scaled residual check will be computed: ||Ax-b||_oo / ( eps * ( || x ||_oo * || A ||_oo + || b ||_oo ) * N ) - The relative machine precision (eps) is taken to be 1.110223e-16 - Computational tests pass if scaled residuals are less than 16.0 Assigning device 1 to process on node tesla2.localdomain rank 1 Assigning device 0 to process on node tesla2.localdomain rank 0 Assigning device 2 to process on node tesla2.localdomain rank 2 Assigning device 3 to process on node tesla2.localdomain rank 3 DTRSM split from environment variable 0.760000 DTRSM split from environment variable 0.760000 DGEMM split from environment variable 0.900000 DGEMM split from environment variable 0.900000 DTRSM split from environment variable 0.760000 DGEMM split from environment variable 0.900000 DTRSM split from environment variable 0.760000 DGEMM split from environment variable 0.900000 ================================================================================ T/V N NB P Q Time Gflops -------------------------------------------------------------------------------- WR00L2L2 16000 1120 2 2 22.40 1.219e+02 -------------------------------------------------------------------------------- ||Ax-b||_oo/(eps*(||A||_oo*||x||_oo+||b||_oo)*N)= 0.0052305 ...... PASSED ================================================================================ Finished 1 tests with the following results: 1 tests completed and passed residual checks, 0 tests completed and failed residual checks, 0 tests skipped because of illegal input values. -------------------------------------------------------------------------------- End of Tests. ================================================================================ ======================== JOB MAP ======================== Data for node: Name: tesla2.localdomain Num procs: 4 Process OMPI jobid: [45109,1] Process rank: 0 Process OMPI jobid: [45109,1] Process rank: 1 Process OMPI jobid: [45109,1] Process rank: 2 Process OMPI jobid: [45109,1] Process rank: 3 ============================================================= ================================================================================ HPLinpack 2.0 -- High-Performance Linpack benchmark -- September 10, 2008 Written by A. Petitet and R. Clint Whaley, Innovative Computing Laboratory, UTK Modified by Piotr Luszczek, Innovative Computing Laboratory, UTK Modified by Julien Langou, University of Colorado Denver ================================================================================ An explanation of the input/output parameters follows: T/V : Wall time / encoded variant. N : The order of the coefficient matrix A. NB : The partitioning blocking factor. P : The number of process rows. Q : The number of process columns. Time : Time in seconds to solve the linear system. Gflops : Rate of execution for solving the linear system. The following parameter values will be used: N : 16000 NB : 1120 PMAP : Row-major process mapping P : 2 Q : 2 PFACT : Right NBMIN : 2 NDIV : 2 RFACT : Right BCAST : 1ring DEPTH : 0 SWAP : Binary-exchange L1 : no-transposed form U : no-transposed form EQUIL : no ALIGN : 8 double precision words -------------------------------------------------------------------------------- - The matrix A is randomly generated for each test. - The following scaled residual check will be computed: ||Ax-b||_oo / ( eps * ( || x ||_oo * || A ||_oo + || b ||_oo ) * N ) - The relative machine precision (eps) is taken to be 1.110223e-16 - Computational tests pass if scaled residuals are less than 16.0 Assigning device 0 to process on node tesla2.localdomain rank 0 Assigning device 2 to process on node tesla2.localdomain rank 2 Assigning device 1 to process on node tesla2.localdomain rank 1 Assigning device 3 to process on node tesla2.localdomain rank 3 DTRSM split from environment variable 0.760000 DTRSM split from environment variable 0.760000 DGEMM split from environment variable 0.900000 DGEMM split from environment variable 0.900000 DTRSM split from environment variable 0.760000 DGEMM split from environment variable 0.900000 DTRSM split from environment variable 0.760000 DGEMM split from environment variable 0.900000 ================================================================================ T/V N NB P Q Time Gflops -------------------------------------------------------------------------------- WR00R2R2 16000 1120 2 2 21.97 1.243e+02 -------------------------------------------------------------------------------- ||Ax-b||_oo/(eps*(||A||_oo*||x||_oo+||b||_oo)*N)= 0.0052305 ...... PASSED ================================================================================ Finished 1 tests with the following results: 1 tests completed and passed residual checks, 0 tests completed and failed residual checks, 0 tests skipped because of illegal input values. -------------------------------------------------------------------------------- End of Tests. ================================================================================ ======================== JOB MAP ======================== Data for node: Name: tesla2.localdomain Num procs: 4 Process OMPI jobid: [49119,1] Process rank: 0 Process OMPI jobid: [49119,1] Process rank: 1 Process OMPI jobid: [49119,1] Process rank: 2 Process OMPI jobid: [49119,1] Process rank: 3 ============================================================= ================================================================================ HPLinpack 2.0 -- High-Performance Linpack benchmark -- September 10, 2008 Written by A. Petitet and R. Clint Whaley, Innovative Computing Laboratory, UTK Modified by Piotr Luszczek, Innovative Computing Laboratory, UTK Modified by Julien Langou, University of Colorado Denver ================================================================================ An explanation of the input/output parameters follows: T/V : Wall time / encoded variant. N : The order of the coefficient matrix A. NB : The partitioning blocking factor. P : The number of process rows. Q : The number of process columns. Time : Time in seconds to solve the linear system. Gflops : Rate of execution for solving the linear system. The following parameter values will be used: N : 16000 NB : 1120 PMAP : Row-major process mapping P : 2 Q : 2 PFACT : Right NBMIN : 2 NDIV : 2 RFACT : Right BCAST : 1ringM DEPTH : 1 SWAP : Binary-exchange L1 : no-transposed form U : no-transposed form EQUIL : no ALIGN : 8 double precision words -------------------------------------------------------------------------------- - The matrix A is randomly generated for each test. - The following scaled residual check will be computed: ||Ax-b||_oo / ( eps * ( || x ||_oo * || A ||_oo + || b ||_oo ) * N ) - The relative machine precision (eps) is taken to be 1.110223e-16 - Computational tests pass if scaled residuals are less than 16.0 Assigning device 0 to process on node tesla2.localdomain rank 0 Assigning device 2 to process on node tesla2.localdomain rank 2 Assigning device 1 to process on node tesla2.localdomain rank 1 Assigning device 3 to process on node tesla2.localdomain rank 3 DTRSM split from environment variable 0.760000 DTRSM split from environment variable 0.760000 DGEMM split from environment variable 0.900000 DGEMM split from environment variable 0.900000 DTRSM split from environment variable 0.760000 DGEMM split from environment variable 0.900000 DTRSM split from environment variable 0.760000 DGEMM split from environment variable 0.900000 ================================================================================ T/V N NB P Q Time Gflops -------------------------------------------------------------------------------- WR11R2R2 16000 1120 2 2 21.53 1.268e+02 -------------------------------------------------------------------------------- ||Ax-b||_oo/(eps*(||A||_oo*||x||_oo+||b||_oo)*N)= 0.0054752 ...... PASSED ================================================================================ Finished 1 tests with the following results: 1 tests completed and passed residual checks, 0 tests completed and failed residual checks, 0 tests skipped because of illegal input values. -------------------------------------------------------------------------------- End of Tests. ================================================================================ ======================== JOB MAP ======================== Data for node: Name: tesla2.localdomain Num procs: 4 Process OMPI jobid: [49032,1] Process rank: 0 Process OMPI jobid: [49032,1] Process rank: 1 Process OMPI jobid: [49032,1] Process rank: 2 Process OMPI jobid: [49032,1] Process rank: 3 ============================================================= ================================================================================ HPLinpack 2.0 -- High-Performance Linpack benchmark -- September 10, 2008 Written by A. Petitet and R. Clint Whaley, Innovative Computing Laboratory, UTK Modified by Piotr Luszczek, Innovative Computing Laboratory, UTK Modified by Julien Langou, University of Colorado Denver ================================================================================ An explanation of the input/output parameters follows: T/V : Wall time / encoded variant. N : The order of the coefficient matrix A. NB : The partitioning blocking factor. P : The number of process rows. Q : The number of process columns. Time : Time in seconds to solve the linear system. Gflops : Rate of execution for solving the linear system. The following parameter values will be used: N : 60000 NB : 1120 PMAP : Row-major process mapping P : 2 Q : 2 PFACT : Right NBMIN : 2 NDIV : 2 RFACT : Right BCAST : 1ringM DEPTH : 1 SWAP : Binary-exchange L1 : no-transposed form U : no-transposed form EQUIL : no ALIGN : 8 double precision words -------------------------------------------------------------------------------- - The matrix A is randomly generated for each test. - The following scaled residual check will be computed: ||Ax-b||_oo / ( eps * ( || x ||_oo * || A ||_oo + || b ||_oo ) * N ) - The relative machine precision (eps) is taken to be 1.110223e-16 - Computational tests pass if scaled residuals are less than 16.0 Assigning device 0 to process on node tesla2.localdomain rank 0 Assigning device 2 to process on node tesla2.localdomain rank 2 Assigning device 1 to process on node tesla2.localdomain rank 1 Assigning device 3 to process on node tesla2.localdomain rank 3 DTRSM split from environment variable 0.760000 DTRSM split from environment variable 0.760000 DGEMM split from environment variable 0.900000 DGEMM split from environment variable 0.900000 DTRSM split from environment variable 0.760000 DGEMM split from environment variable 0.900000 DTRSM split from environment variable 0.760000 DGEMM split from environment variable 0.900000 ================================================================================ T/V N NB P Q Time Gflops -------------------------------------------------------------------------------- WR11R2R2 60000 1120 2 2 624.74 2.305e+02 -------------------------------------------------------------------------------- ||Ax-b||_oo/(eps*(||A||_oo*||x||_oo+||b||_oo)*N)= 0.0044561 ...... PASSED ================================================================================ Finished 1 tests with the following results: 1 tests completed and passed residual checks, 0 tests completed and failed residual checks, 0 tests skipped because of illegal input values. -------------------------------------------------------------------------------- End of Tests. ================================================================================ ======================== JOB MAP ======================== Data for node: Name: node-13 Num procs: 8 Process OMPI jobid: [28556,1] Process rank: 0 Process OMPI jobid: [28556,1] Process rank: 4 Process OMPI jobid: [28556,1] Process rank: 8 Process OMPI jobid: [28556,1] Process rank: 12 Process OMPI jobid: [28556,1] Process rank: 16 Process OMPI jobid: [28556,1] Process rank: 20 Process OMPI jobid: [28556,1] Process rank: 24 Process OMPI jobid: [28556,1] Process rank: 28 Data for node: Name: node-14 Num procs: 8 Process OMPI jobid: [28556,1] Process rank: 1 Process OMPI jobid: [28556,1] Process rank: 5 Process OMPI jobid: [28556,1] Process rank: 9 Process OMPI jobid: [28556,1] Process rank: 13 Process OMPI jobid: [28556,1] Process rank: 17 Process OMPI jobid: [28556,1] Process rank: 21 Process OMPI jobid: [28556,1] Process rank: 25 Process OMPI jobid: [28556,1] Process rank: 29 Data for node: Name: node-15 Num procs: 8 Process OMPI jobid: [28556,1] Process rank: 2 Process OMPI jobid: [28556,1] Process rank: 6 Process OMPI jobid: [28556,1] Process rank: 10 Process OMPI jobid: [28556,1] Process rank: 14 Process OMPI jobid: [28556,1] Process rank: 18 Process OMPI jobid: [28556,1] Process rank: 22 Process OMPI jobid: [28556,1] Process rank: 26 Process OMPI jobid: [28556,1] Process rank: 30 Data for node: Name: node-16 Num procs: 8 Process OMPI jobid: [28556,1] Process rank: 3 Process OMPI jobid: [28556,1] Process rank: 7 Process OMPI jobid: [28556,1] Process rank: 11 Process OMPI jobid: [28556,1] Process rank: 15 Process OMPI jobid: [28556,1] Process rank: 19 Process OMPI jobid: [28556,1] Process rank: 23 Process OMPI jobid: [28556,1] Process rank: 27 Process OMPI jobid: [28556,1] Process rank: 31 ============================================================= bash: /share/mpi/openmpi-1.4.2-gcc/bin/orted: No such file or directory -------------------------------------------------------------------------- A daemon (pid 30505) died unexpectedly with status 127 while attempting to launch so we are aborting. There may be more information reported by the environment (see above). This may be because the daemon was unable to find all the needed shared libraries on the remote node. You may set your LD_LIBRARY_PATH to have the location of the shared libraries on the remote nodes and this will automatically be forwarded to the remote nodes. -------------------------------------------------------------------------- -------------------------------------------------------------------------- mpirun noticed that the job aborted, but has no info as to the process that caused that situation. -------------------------------------------------------------------------- mpirun: clean termination accomplished ======================== JOB MAP ======================== Data for node: Name: node-13 Num procs: 8 Process OMPI jobid: [28504,1] Process rank: 0 Process OMPI jobid: [28504,1] Process rank: 4 Process OMPI jobid: [28504,1] Process rank: 8 Process OMPI jobid: [28504,1] Process rank: 12 Process OMPI jobid: [28504,1] Process rank: 16 Process OMPI jobid: [28504,1] Process rank: 20 Process OMPI jobid: [28504,1] Process rank: 24 Process OMPI jobid: [28504,1] Process rank: 28 Data for node: Name: node-14 Num procs: 8 Process OMPI jobid: [28504,1] Process rank: 1 Process OMPI jobid: [28504,1] Process rank: 5 Process OMPI jobid: [28504,1] Process rank: 9 Process OMPI jobid: [28504,1] Process rank: 13 Process OMPI jobid: [28504,1] Process rank: 17 Process OMPI jobid: [28504,1] Process rank: 21 Process OMPI jobid: [28504,1] Process rank: 25 Process OMPI jobid: [28504,1] Process rank: 29 Data for node: Name: node-15 Num procs: 8 Process OMPI jobid: [28504,1] Process rank: 2 Process OMPI jobid: [28504,1] Process rank: 6 Process OMPI jobid: [28504,1] Process rank: 10 Process OMPI jobid: [28504,1] Process rank: 14 Process OMPI jobid: [28504,1] Process rank: 18 Process OMPI jobid: [28504,1] Process rank: 22 Process OMPI jobid: [28504,1] Process rank: 26 Process OMPI jobid: [28504,1] Process rank: 30 Data for node: Name: node-16 Num procs: 8 Process OMPI jobid: [28504,1] Process rank: 3 Process OMPI jobid: [28504,1] Process rank: 7 Process OMPI jobid: [28504,1] Process rank: 11 Process OMPI jobid: [28504,1] Process rank: 15 Process OMPI jobid: [28504,1] Process rank: 19 Process OMPI jobid: [28504,1] Process rank: 23 Process OMPI jobid: [28504,1] Process rank: 27 Process OMPI jobid: [28504,1] Process rank: 31 ============================================================= bash: /share/mpi/openmpi-1.4.2-gcc/bin/orted: No such file or directory -------------------------------------------------------------------------- A daemon (pid 30709) died unexpectedly with status 127 while attempting to launch so we are aborting. There may be more information reported by the environment (see above). This may be because the daemon was unable to find all the needed shared libraries on the remote node. You may set your LD_LIBRARY_PATH to have the location of the shared libraries on the remote nodes and this will automatically be forwarded to the remote nodes. -------------------------------------------------------------------------- -------------------------------------------------------------------------- mpirun noticed that the job aborted, but has no info as to the process that caused that situation. -------------------------------------------------------------------------- mpirun: clean termination accomplished ======================== JOB MAP ======================== Data for node: Name: node-13 Num procs: 8 Process OMPI jobid: [12797,1] Process rank: 0 Process OMPI jobid: [12797,1] Process rank: 4 Process OMPI jobid: [12797,1] Process rank: 8 Process OMPI jobid: [12797,1] Process rank: 12 Process OMPI jobid: [12797,1] Process rank: 16 Process OMPI jobid: [12797,1] Process rank: 20 Process OMPI jobid: [12797,1] Process rank: 24 Process OMPI jobid: [12797,1] Process rank: 28 Data for node: Name: node-14 Num procs: 8 Process OMPI jobid: [12797,1] Process rank: 1 Process OMPI jobid: [12797,1] Process rank: 5 Process OMPI jobid: [12797,1] Process rank: 9 Process OMPI jobid: [12797,1] Process rank: 13 Process OMPI jobid: [12797,1] Process rank: 17 Process OMPI jobid: [12797,1] Process rank: 21 Process OMPI jobid: [12797,1] Process rank: 25 Process OMPI jobid: [12797,1] Process rank: 29 Data for node: Name: node-15 Num procs: 8 Process OMPI jobid: [12797,1] Process rank: 2 Process OMPI jobid: [12797,1] Process rank: 6 Process OMPI jobid: [12797,1] Process rank: 10 Process OMPI jobid: [12797,1] Process rank: 14 Process OMPI jobid: [12797,1] Process rank: 18 Process OMPI jobid: [12797,1] Process rank: 22 Process OMPI jobid: [12797,1] Process rank: 26 Process OMPI jobid: [12797,1] Process rank: 30 Data for node: Name: node-16 Num procs: 8 Process OMPI jobid: [12797,1] Process rank: 3 Process OMPI jobid: [12797,1] Process rank: 7 Process OMPI jobid: [12797,1] Process rank: 11 Process OMPI jobid: [12797,1] Process rank: 15 Process OMPI jobid: [12797,1] Process rank: 19 Process OMPI jobid: [12797,1] Process rank: 23 Process OMPI jobid: [12797,1] Process rank: 27 Process OMPI jobid: [12797,1] Process rank: 31 ============================================================= Warning: Permanently added 'node-15,192.168.0.15' (RSA) to the list of known hosts. Warning: Permanently added 'node-16,192.168.0.16' (RSA) to the list of known hosts. Warning: Permanently added 'node-14,192.168.0.14' (RSA) to the list of known hosts. bash: /share/mpi/openmpi-1.4.2-gcc/bin/orted: No such file or directory -------------------------------------------------------------------------- A daemon (pid 3496) died unexpectedly with status 127 while attempting to launch so we are aborting. There may be more information reported by the environment (see above). This may be because the daemon was unable to find all the needed shared libraries on the remote node. You may set your LD_LIBRARY_PATH to have the location of the shared libraries on the remote nodes and this will automatically be forwarded to the remote nodes. -------------------------------------------------------------------------- -------------------------------------------------------------------------- mpirun noticed that the job aborted, but has no info as to the process that caused that situation. -------------------------------------------------------------------------- -------------------------------------------------------------------------- mpirun was unable to cleanly terminate the daemons on the nodes shown below. Additional manual cleanup may be required - please refer to the "orte-clean" tool for assistance. -------------------------------------------------------------------------- node-14 - daemon did not report back when launched node-16 - daemon did not report back when launched Permission denied, please try again. Permission denied, please try again.