API tools faq
paste
Advertisement
Guest User

Untitled

a guest
Aug 6th, 2017
157
0
Never
Add comment
Not a member of Pastebin yet? Sign Up, it unlocks many cool features!
text 40.73 KB | None | 0 0
raw download clone embed print report
  1. root@espressobin:/usr/local/src/StabilityTester# while true ; do ./xhpl64 ; done
  2. ================================================================================
  3. HPLinpack 2.1 -- High-Performance Linpack benchmark -- October 26, 2012
  4. Written by A. Petitet and R. Clint Whaley, Innovative Computing Laboratory, UTK
  5. Modified by Piotr Luszczek, Innovative Computing Laboratory, UTK
  6. Modified by Julien Langou, University of Colorado Denver
  7. ================================================================================
  8.  
  9. An explanation of the input/output parameters follows:
  10. T/V : Wall time / encoded variant.
  11. N : The order of the coefficient matrix A.
  12. NB : The partitioning blocking factor.
  13. P : The number of process rows.
  14. Q : The number of process columns.
  15. Time : Time in seconds to solve the linear system.
  16. Gflops : Rate of execution for solving the linear system.
  17.  
  18. The following parameter values will be used:
  19.  
  20. N : 4096
  21. NB : 256
  22. PMAP : Row-major process mapping
  23. P : 1
  24. Q : 1
  25. PFACT : Left
  26. NBMIN : 2
  27. NDIV : 2
  28. RFACT : Right
  29. BCAST : 2ring
  30. DEPTH : 0
  31. SWAP : Mix (threshold = 64)
  32. L1 : transposed form
  33. U : transposed form
  34. EQUIL : yes
  35. ALIGN : 8 double precision words
  36.  
  37. --------------------------------------------------------------------------------
  38.  
  39. - The matrix A is randomly generated for each test.
  40. - The following scaled residual check will be computed:
  41. ||Ax-b||_oo / ( eps * ( || x ||_oo * || A ||_oo + || b ||_oo ) * N )
  42. - The relative machine precision (eps) is taken to be 1.110223e-16
  43. - Computational tests pass if scaled residuals are less than 0.0
  44.  
  45. ================================================================================
  46. T/V N NB P Q Time Gflops
  47. --------------------------------------------------------------------------------
  48. WR02R2L2 4096 256 1 1 31.09 1.474e+00
  49. HPL_pdgesv() start time Sun Aug 6 13:19:42 2017
  50.  
  51. HPL_pdgesv() end time Sun Aug 6 13:20:14 2017
  52.  
  53. --------------------------------------------------------------------------------
  54. ||Ax-b||_oo/(eps*(||A||_oo*||x||_oo+||b||_oo)*N)= 0.0048034 ...... PASSED
  55. ================================================================================
  56.  
  57. Finished 1 tests with the following results:
  58. 1 tests completed and passed residual checks,
  59. 0 tests completed and failed residual checks,
  60. 0 tests skipped because of illegal input values.
  61. --------------------------------------------------------------------------------
  62.  
  63. End of Tests.
  64. ================================================================================
  65. ================================================================================
  66. HPLinpack 2.1 -- High-Performance Linpack benchmark -- October 26, 2012
  67. Written by A. Petitet and R. Clint Whaley, Innovative Computing Laboratory, UTK
  68. Modified by Piotr Luszczek, Innovative Computing Laboratory, UTK
  69. Modified by Julien Langou, University of Colorado Denver
  70. ================================================================================
  71.  
  72. An explanation of the input/output parameters follows:
  73. T/V : Wall time / encoded variant.
  74. N : The order of the coefficient matrix A.
  75. NB : The partitioning blocking factor.
  76. P : The number of process rows.
  77. Q : The number of process columns.
  78. Time : Time in seconds to solve the linear system.
  79. Gflops : Rate of execution for solving the linear system.
  80.  
  81. The following parameter values will be used:
  82.  
  83. N : 4096
  84. NB : 256
  85. PMAP : Row-major process mapping
  86. P : 1
  87. Q : 1
  88. PFACT : Left
  89. NBMIN : 2
  90. NDIV : 2
  91. RFACT : Right
  92. BCAST : 2ring
  93. DEPTH : 0
  94. SWAP : Mix (threshold = 64)
  95. L1 : transposed form
  96. U : transposed form
  97. EQUIL : yes
  98. ALIGN : 8 double precision words
  99.  
  100. --------------------------------------------------------------------------------
  101.  
  102. - The matrix A is randomly generated for each test.
  103. - The following scaled residual check will be computed:
  104. ||Ax-b||_oo / ( eps * ( || x ||_oo * || A ||_oo + || b ||_oo ) * N )
  105. - The relative machine precision (eps) is taken to be 1.110223e-16
  106. - Computational tests pass if scaled residuals are less than 0.0
  107.  
  108. ================================================================================
  109. T/V N NB P Q Time Gflops
  110. --------------------------------------------------------------------------------
  111. WR02R2L2 4096 256 1 1 30.77 1.490e+00
  112. HPL_pdgesv() start time Sun Aug 6 13:20:19 2017
  113.  
  114. HPL_pdgesv() end time Sun Aug 6 13:20:50 2017
  115.  
  116. --------------------------------------------------------------------------------
  117. ||Ax-b||_oo/(eps*(||A||_oo*||x||_oo+||b||_oo)*N)= 0.0048034 ...... PASSED
  118. ================================================================================
  119.  
  120. Finished 1 tests with the following results:
  121. 1 tests completed and passed residual checks,
  122. 0 tests completed and failed residual checks,
  123. 0 tests skipped because of illegal input values.
  124. --------------------------------------------------------------------------------
  125.  
  126. End of Tests.
  127. ================================================================================
  128. ================================================================================
  129. HPLinpack 2.1 -- High-Performance Linpack benchmark -- October 26, 2012
  130. Written by A. Petitet and R. Clint Whaley, Innovative Computing Laboratory, UTK
  131. Modified by Piotr Luszczek, Innovative Computing Laboratory, UTK
  132. Modified by Julien Langou, University of Colorado Denver
  133. ================================================================================
  134.  
  135. An explanation of the input/output parameters follows:
  136. T/V : Wall time / encoded variant.
  137. N : The order of the coefficient matrix A.
  138. NB : The partitioning blocking factor.
  139. P : The number of process rows.
  140. Q : The number of process columns.
  141. Time : Time in seconds to solve the linear system.
  142. Gflops : Rate of execution for solving the linear system.
  143.  
  144. The following parameter values will be used:
  145.  
  146. N : 4096
  147. NB : 256
  148. PMAP : Row-major process mapping
  149. P : 1
  150. Q : 1
  151. PFACT : Left
  152. NBMIN : 2
  153. NDIV : 2
  154. RFACT : Right
  155. BCAST : 2ring
  156. DEPTH : 0
  157. SWAP : Mix (threshold = 64)
  158. L1 : transposed form
  159. U : transposed form
  160. EQUIL : yes
  161. ALIGN : 8 double precision words
  162.  
  163. --------------------------------------------------------------------------------
  164.  
  165. - The matrix A is randomly generated for each test.
  166. - The following scaled residual check will be computed:
  167. ||Ax-b||_oo / ( eps * ( || x ||_oo * || A ||_oo + || b ||_oo ) * N )
  168. - The relative machine precision (eps) is taken to be 1.110223e-16
  169. - Computational tests pass if scaled residuals are less than 0.0
  170.  
  171. ================================================================================
  172. T/V N NB P Q Time Gflops
  173. --------------------------------------------------------------------------------
  174. WR02R2L2 4096 256 1 1 30.89 1.484e+00
  175. HPL_pdgesv() start time Sun Aug 6 13:20:55 2017
  176.  
  177. HPL_pdgesv() end time Sun Aug 6 13:21:26 2017
  178.  
  179. --------------------------------------------------------------------------------
  180. ||Ax-b||_oo/(eps*(||A||_oo*||x||_oo+||b||_oo)*N)= 0.0048034 ...... PASSED
  181. ================================================================================
  182.  
  183. Finished 1 tests with the following results:
  184. 1 tests completed and passed residual checks,
  185. 0 tests completed and failed residual checks,
  186. 0 tests skipped because of illegal input values.
  187. --------------------------------------------------------------------------------
  188.  
  189. End of Tests.
  190. ================================================================================
  191. ================================================================================
  192. HPLinpack 2.1 -- High-Performance Linpack benchmark -- October 26, 2012
  193. Written by A. Petitet and R. Clint Whaley, Innovative Computing Laboratory, UTK
  194. Modified by Piotr Luszczek, Innovative Computing Laboratory, UTK
  195. Modified by Julien Langou, University of Colorado Denver
  196. ================================================================================
  197.  
  198. An explanation of the input/output parameters follows:
  199. T/V : Wall time / encoded variant.
  200. N : The order of the coefficient matrix A.
  201. NB : The partitioning blocking factor.
  202. P : The number of process rows.
  203. Q : The number of process columns.
  204. Time : Time in seconds to solve the linear system.
  205. Gflops : Rate of execution for solving the linear system.
  206.  
  207. The following parameter values will be used:
  208.  
  209. N : 4096
  210. NB : 256
  211. PMAP : Row-major process mapping
  212. P : 1
  213. Q : 1
  214. PFACT : Left
  215. NBMIN : 2
  216. NDIV : 2
  217. RFACT : Right
  218. BCAST : 2ring
  219. DEPTH : 0
  220. SWAP : Mix (threshold = 64)
  221. L1 : transposed form
  222. U : transposed form
  223. EQUIL : yes
  224. ALIGN : 8 double precision words
  225.  
  226. --------------------------------------------------------------------------------
  227.  
  228. - The matrix A is randomly generated for each test.
  229. - The following scaled residual check will be computed:
  230. ||Ax-b||_oo / ( eps * ( || x ||_oo * || A ||_oo + || b ||_oo ) * N )
  231. - The relative machine precision (eps) is taken to be 1.110223e-16
  232. - Computational tests pass if scaled residuals are less than 0.0
  233.  
  234. ================================================================================
  235. T/V N NB P Q Time Gflops
  236. --------------------------------------------------------------------------------
  237. WR02R2L2 4096 256 1 1 31.01 1.478e+00
  238. HPL_pdgesv() start time Sun Aug 6 13:21:31 2017
  239.  
  240. HPL_pdgesv() end time Sun Aug 6 13:22:02 2017
  241.  
  242. --------------------------------------------------------------------------------
  243. ||Ax-b||_oo/(eps*(||A||_oo*||x||_oo+||b||_oo)*N)= 0.0048034 ...... PASSED
  244. ================================================================================
  245.  
  246. Finished 1 tests with the following results:
  247. 1 tests completed and passed residual checks,
  248. 0 tests completed and failed residual checks,
  249. 0 tests skipped because of illegal input values.
  250. --------------------------------------------------------------------------------
  251.  
  252. End of Tests.
  253. ================================================================================
  254. ================================================================================
  255. HPLinpack 2.1 -- High-Performance Linpack benchmark -- October 26, 2012
  256. Written by A. Petitet and R. Clint Whaley, Innovative Computing Laboratory, UTK
  257. Modified by Piotr Luszczek, Innovative Computing Laboratory, UTK
  258. Modified by Julien Langou, University of Colorado Denver
  259. ================================================================================
  260.  
  261. An explanation of the input/output parameters follows:
  262. T/V : Wall time / encoded variant.
  263. N : The order of the coefficient matrix A.
  264. NB : The partitioning blocking factor.
  265. P : The number of process rows.
  266. Q : The number of process columns.
  267. Time : Time in seconds to solve the linear system.
  268. Gflops : Rate of execution for solving the linear system.
  269.  
  270. The following parameter values will be used:
  271.  
  272. N : 4096
  273. NB : 256
  274. PMAP : Row-major process mapping
  275. P : 1
  276. Q : 1
  277. PFACT : Left
  278. NBMIN : 2
  279. NDIV : 2
  280. RFACT : Right
  281. BCAST : 2ring
  282. DEPTH : 0
  283. SWAP : Mix (threshold = 64)
  284. L1 : transposed form
  285. U : transposed form
  286. EQUIL : yes
  287. ALIGN : 8 double precision words
  288.  
  289. --------------------------------------------------------------------------------
  290.  
  291. - The matrix A is randomly generated for each test.
  292. - The following scaled residual check will be computed:
  293. ||Ax-b||_oo / ( eps * ( || x ||_oo * || A ||_oo + || b ||_oo ) * N )
  294. - The relative machine precision (eps) is taken to be 1.110223e-16
  295. - Computational tests pass if scaled residuals are less than 0.0
  296.  
  297. ================================================================================
  298. T/V N NB P Q Time Gflops
  299. --------------------------------------------------------------------------------
  300. WR02R2L2 4096 256 1 1 30.85 1.486e+00
  301. HPL_pdgesv() start time Sun Aug 6 13:22:07 2017
  302.  
  303. HPL_pdgesv() end time Sun Aug 6 13:22:38 2017
  304.  
  305. --------------------------------------------------------------------------------
  306. ||Ax-b||_oo/(eps*(||A||_oo*||x||_oo+||b||_oo)*N)= 0.0048034 ...... PASSED
  307. ================================================================================
  308.  
  309. Finished 1 tests with the following results:
  310. 1 tests completed and passed residual checks,
  311. 0 tests completed and failed residual checks,
  312. 0 tests skipped because of illegal input values.
  313. --------------------------------------------------------------------------------
  314.  
  315. End of Tests.
  316. ================================================================================
  317. ================================================================================
  318. HPLinpack 2.1 -- High-Performance Linpack benchmark -- October 26, 2012
  319. Written by A. Petitet and R. Clint Whaley, Innovative Computing Laboratory, UTK
  320. Modified by Piotr Luszczek, Innovative Computing Laboratory, UTK
  321. Modified by Julien Langou, University of Colorado Denver
  322. ================================================================================
  323.  
  324. An explanation of the input/output parameters follows:
  325. T/V : Wall time / encoded variant.
  326. N : The order of the coefficient matrix A.
  327. NB : The partitioning blocking factor.
  328. P : The number of process rows.
  329. Q : The number of process columns.
  330. Time : Time in seconds to solve the linear system.
  331. Gflops : Rate of execution for solving the linear system.
  332.  
  333. The following parameter values will be used:
  334.  
  335. N : 4096
  336. NB : 256
  337. PMAP : Row-major process mapping
  338. P : 1
  339. Q : 1
  340. PFACT : Left
  341. NBMIN : 2
  342. NDIV : 2
  343. RFACT : Right
  344. BCAST : 2ring
  345. DEPTH : 0
  346. SWAP : Mix (threshold = 64)
  347. L1 : transposed form
  348. U : transposed form
  349. EQUIL : yes
  350. ALIGN : 8 double precision words
  351.  
  352. --------------------------------------------------------------------------------
  353.  
  354. - The matrix A is randomly generated for each test.
  355. - The following scaled residual check will be computed:
  356. ||Ax-b||_oo / ( eps * ( || x ||_oo * || A ||_oo + || b ||_oo ) * N )
  357. - The relative machine precision (eps) is taken to be 1.110223e-16
  358. - Computational tests pass if scaled residuals are less than 0.0
  359.  
  360. ================================================================================
  361. T/V N NB P Q Time Gflops
  362. --------------------------------------------------------------------------------
  363. WR02R2L2 4096 256 1 1 30.91 1.483e+00
  364. HPL_pdgesv() start time Sun Aug 6 13:22:43 2017
  365.  
  366. HPL_pdgesv() end time Sun Aug 6 13:23:14 2017
  367.  
  368. --------------------------------------------------------------------------------
  369. ||Ax-b||_oo/(eps*(||A||_oo*||x||_oo+||b||_oo)*N)= 0.0048034 ...... PASSED
  370. ================================================================================
  371.  
  372. Finished 1 tests with the following results:
  373. 1 tests completed and passed residual checks,
  374. 0 tests completed and failed residual checks,
  375. 0 tests skipped because of illegal input values.
  376. --------------------------------------------------------------------------------
  377.  
  378. End of Tests.
  379. ================================================================================
  380. ================================================================================
  381. HPLinpack 2.1 -- High-Performance Linpack benchmark -- October 26, 2012
  382. Written by A. Petitet and R. Clint Whaley, Innovative Computing Laboratory, UTK
  383. Modified by Piotr Luszczek, Innovative Computing Laboratory, UTK
  384. Modified by Julien Langou, University of Colorado Denver
  385. ================================================================================
  386.  
  387. An explanation of the input/output parameters follows:
  388. T/V : Wall time / encoded variant.
  389. N : The order of the coefficient matrix A.
  390. NB : The partitioning blocking factor.
  391. P : The number of process rows.
  392. Q : The number of process columns.
  393. Time : Time in seconds to solve the linear system.
  394. Gflops : Rate of execution for solving the linear system.
  395.  
  396. The following parameter values will be used:
  397.  
  398. N : 4096
  399. NB : 256
  400. PMAP : Row-major process mapping
  401. P : 1
  402. Q : 1
  403. PFACT : Left
  404. NBMIN : 2
  405. NDIV : 2
  406. RFACT : Right
  407. BCAST : 2ring
  408. DEPTH : 0
  409. SWAP : Mix (threshold = 64)
  410. L1 : transposed form
  411. U : transposed form
  412. EQUIL : yes
  413. ALIGN : 8 double precision words
  414.  
  415. --------------------------------------------------------------------------------
  416.  
  417. - The matrix A is randomly generated for each test.
  418. - The following scaled residual check will be computed:
  419. ||Ax-b||_oo / ( eps * ( || x ||_oo * || A ||_oo + || b ||_oo ) * N )
  420. - The relative machine precision (eps) is taken to be 1.110223e-16
  421. - Computational tests pass if scaled residuals are less than 0.0
  422.  
  423.  
  424.  
  425.  
  426.  
  427.  
  428.  
  429.  
  430. ================================================================================
  431. T/V N NB P Q Time Gflops
  432. --------------------------------------------------------------------------------
  433. WR02R2L2 4096 256 1 1 30.83 1.487e+00
  434. HPL_pdgesv() start time Sun Aug 6 13:23:19 2017
  435.  
  436. HPL_pdgesv() end time Sun Aug 6 13:23:50 2017
  437.  
  438. --------------------------------------------------------------------------------
  439. ||Ax-b||_oo/(eps*(||A||_oo*||x||_oo+||b||_oo)*N)= 0.0048034 ...... PASSED
  440. ================================================================================
  441.  
  442. Finished 1 tests with the following results:
  443. 1 tests completed and passed residual checks,
  444. 0 tests completed and failed residual checks,
  445. 0 tests skipped because of illegal input values.
  446. --------------------------------------------------------------------------------
  447.  
  448. End of Tests.
  449. ================================================================================
  450. ================================================================================
  451. HPLinpack 2.1 -- High-Performance Linpack benchmark -- October 26, 2012
  452. Written by A. Petitet and R. Clint Whaley, Innovative Computing Laboratory, UTK
  453. Modified by Piotr Luszczek, Innovative Computing Laboratory, UTK
  454. Modified by Julien Langou, University of Colorado Denver
  455. ================================================================================
  456.  
  457. An explanation of the input/output parameters follows:
  458. T/V : Wall time / encoded variant.
  459. N : The order of the coefficient matrix A.
  460. NB : The partitioning blocking factor.
  461. P : The number of process rows.
  462. Q : The number of process columns.
  463. Time : Time in seconds to solve the linear system.
  464. Gflops : Rate of execution for solving the linear system.
  465.  
  466. The following parameter values will be used:
  467.  
  468. N : 4096
  469. NB : 256
  470. PMAP : Row-major process mapping
  471. P : 1
  472. Q : 1
  473. PFACT : Left
  474. NBMIN : 2
  475. NDIV : 2
  476. RFACT : Right
  477. BCAST : 2ring
  478. DEPTH : 0
  479. SWAP : Mix (threshold = 64)
  480. L1 : transposed form
  481. U : transposed form
  482. EQUIL : yes
  483. ALIGN : 8 double precision words
  484.  
  485. --------------------------------------------------------------------------------
  486.  
  487. - The matrix A is randomly generated for each test.
  488. - The following scaled residual check will be computed:
  489. ||Ax-b||_oo / ( eps * ( || x ||_oo * || A ||_oo + || b ||_oo ) * N )
  490. - The relative machine precision (eps) is taken to be 1.110223e-16
  491. - Computational tests pass if scaled residuals are less than 0.0
  492.  
  493. ================================================================================
  494. T/V N NB P Q Time Gflops
  495. --------------------------------------------------------------------------------
  496. WR02R2L2 4096 256 1 1 30.84 1.486e+00
  497. HPL_pdgesv() start time Sun Aug 6 13:23:55 2017
  498.  
  499. HPL_pdgesv() end time Sun Aug 6 13:24:26 2017
  500.  
  501. --------------------------------------------------------------------------------
  502. ||Ax-b||_oo/(eps*(||A||_oo*||x||_oo+||b||_oo)*N)= 0.0048034 ...... PASSED
  503. ================================================================================
  504.  
  505. Finished 1 tests with the following results:
  506. 1 tests completed and passed residual checks,
  507. 0 tests completed and failed residual checks,
  508. 0 tests skipped because of illegal input values.
  509. --------------------------------------------------------------------------------
  510.  
  511. End of Tests.
  512. ================================================================================
  513. ================================================================================
  514. HPLinpack 2.1 -- High-Performance Linpack benchmark -- October 26, 2012
  515. Written by A. Petitet and R. Clint Whaley, Innovative Computing Laboratory, UTK
  516. Modified by Piotr Luszczek, Innovative Computing Laboratory, UTK
  517. Modified by Julien Langou, University of Colorado Denver
  518. ================================================================================
  519.  
  520. An explanation of the input/output parameters follows:
  521. T/V : Wall time / encoded variant.
  522. N : The order of the coefficient matrix A.
  523. NB : The partitioning blocking factor.
  524. P : The number of process rows.
  525. Q : The number of process columns.
  526. Time : Time in seconds to solve the linear system.
  527. Gflops : Rate of execution for solving the linear system.
  528.  
  529. The following parameter values will be used:
  530.  
  531. N : 4096
  532. NB : 256
  533. PMAP : Row-major process mapping
  534. P : 1
  535. Q : 1
  536. PFACT : Left
  537. NBMIN : 2
  538. NDIV : 2
  539. RFACT : Right
  540. BCAST : 2ring
  541. DEPTH : 0
  542. SWAP : Mix (threshold = 64)
  543. L1 : transposed form
  544. U : transposed form
  545. EQUIL : yes
  546. ALIGN : 8 double precision words
  547.  
  548. --------------------------------------------------------------------------------
  549.  
  550. - The matrix A is randomly generated for each test.
  551. - The following scaled residual check will be computed:
  552. ||Ax-b||_oo / ( eps * ( || x ||_oo * || A ||_oo + || b ||_oo ) * N )
  553. - The relative machine precision (eps) is taken to be 1.110223e-16
  554. - Computational tests pass if scaled residuals are less than 0.0
  555.  
  556. ================================================================================
  557. T/V N NB P Q Time Gflops
  558. --------------------------------------------------------------------------------
  559. WR02R2L2 4096 256 1 1 30.90 1.483e+00
  560. HPL_pdgesv() start time Sun Aug 6 13:24:31 2017
  561.  
  562. HPL_pdgesv() end time Sun Aug 6 13:25:02 2017
  563.  
  564. --------------------------------------------------------------------------------
  565. ||Ax-b||_oo/(eps*(||A||_oo*||x||_oo+||b||_oo)*N)= 0.0048034 ...... PASSED
  566. ================================================================================
  567.  
  568. Finished 1 tests with the following results:
  569. 1 tests completed and passed residual checks,
  570. 0 tests completed and failed residual checks,
  571. 0 tests skipped because of illegal input values.
  572. --------------------------------------------------------------------------------
  573.  
  574. End of Tests.
  575. ================================================================================
  576. ================================================================================
  577. HPLinpack 2.1 -- High-Performance Linpack benchmark -- October 26, 2012
  578. Written by A. Petitet and R. Clint Whaley, Innovative Computing Laboratory, UTK
  579. Modified by Piotr Luszczek, Innovative Computing Laboratory, UTK
  580. Modified by Julien Langou, University of Colorado Denver
  581. ================================================================================
  582.  
  583. An explanation of the input/output parameters follows:
  584. T/V : Wall time / encoded variant.
  585. N : The order of the coefficient matrix A.
  586. NB : The partitioning blocking factor.
  587. P : The number of process rows.
  588. Q : The number of process columns.
  589. Time : Time in seconds to solve the linear system.
  590. Gflops : Rate of execution for solving the linear system.
  591.  
  592. The following parameter values will be used:
  593.  
  594. N : 4096
  595. NB : 256
  596. PMAP : Row-major process mapping
  597. P : 1
  598. Q : 1
  599. PFACT : Left
  600. NBMIN : 2
  601. NDIV : 2
  602. RFACT : Right
  603. BCAST : 2ring
  604. DEPTH : 0
  605. SWAP : Mix (threshold = 64)
  606. L1 : transposed form
  607. U : transposed form
  608. EQUIL : yes
  609. ALIGN : 8 double precision words
  610.  
  611. --------------------------------------------------------------------------------
  612.  
  613. - The matrix A is randomly generated for each test.
  614. - The following scaled residual check will be computed:
  615. ||Ax-b||_oo / ( eps * ( || x ||_oo * || A ||_oo + || b ||_oo ) * N )
  616. - The relative machine precision (eps) is taken to be 1.110223e-16
  617. - Computational tests pass if scaled residuals are less than 0.0
  618.  
  619. ================================================================================
  620. T/V N NB P Q Time Gflops
  621. --------------------------------------------------------------------------------
  622. WR02R2L2 4096 256 1 1 31.08 1.475e+00
  623. HPL_pdgesv() start time Sun Aug 6 13:25:08 2017
  624.  
  625. HPL_pdgesv() end time Sun Aug 6 13:25:39 2017
  626.  
  627. --------------------------------------------------------------------------------
  628. ||Ax-b||_oo/(eps*(||A||_oo*||x||_oo+||b||_oo)*N)= 0.0048034 ...... PASSED
  629. ================================================================================
  630.  
  631. Finished 1 tests with the following results:
  632. 1 tests completed and passed residual checks,
  633. 0 tests completed and failed residual checks,
  634. 0 tests skipped because of illegal input values.
  635. --------------------------------------------------------------------------------
  636.  
  637. End of Tests.
  638. ================================================================================
  639. ================================================================================
  640. HPLinpack 2.1 -- High-Performance Linpack benchmark -- October 26, 2012
  641. Written by A. Petitet and R. Clint Whaley, Innovative Computing Laboratory, UTK
  642. Modified by Piotr Luszczek, Innovative Computing Laboratory, UTK
  643. Modified by Julien Langou, University of Colorado Denver
  644. ================================================================================
  645.  
  646. An explanation of the input/output parameters follows:
  647. T/V : Wall time / encoded variant.
  648. N : The order of the coefficient matrix A.
  649. NB : The partitioning blocking factor.
  650. P : The number of process rows.
  651. Q : The number of process columns.
  652. Time : Time in seconds to solve the linear system.
  653. Gflops : Rate of execution for solving the linear system.
  654.  
  655. The following parameter values will be used:
  656.  
  657. N : 4096
  658. NB : 256
  659. PMAP : Row-major process mapping
  660. P : 1
  661. Q : 1
  662. PFACT : Left
  663. NBMIN : 2
  664. NDIV : 2
  665. RFACT : Right
  666. BCAST : 2ring
  667. DEPTH : 0
  668. SWAP : Mix (threshold = 64)
  669. L1 : transposed form
  670. U : transposed form
  671. EQUIL : yes
  672. ALIGN : 8 double precision words
  673.  
  674. --------------------------------------------------------------------------------
  675.  
  676. - The matrix A is randomly generated for each test.
  677. - The following scaled residual check will be computed:
  678. ||Ax-b||_oo / ( eps * ( || x ||_oo * || A ||_oo + || b ||_oo ) * N )
  679. - The relative machine precision (eps) is taken to be 1.110223e-16
  680. - Computational tests pass if scaled residuals are less than 0.0
  681.  
  682. ================================================================================
  683. T/V N NB P Q Time Gflops
  684. --------------------------------------------------------------------------------
  685. WR02R2L2 4096 256 1 1 30.90 1.483e+00
  686. HPL_pdgesv() start time Sun Aug 6 13:25:44 2017
  687.  
  688. HPL_pdgesv() end time Sun Aug 6 13:26:15 2017
  689.  
  690. --------------------------------------------------------------------------------
  691. ||Ax-b||_oo/(eps*(||A||_oo*||x||_oo+||b||_oo)*N)= 0.0048034 ...... PASSED
  692. ================================================================================
  693.  
  694. Finished 1 tests with the following results:
  695. 1 tests completed and passed residual checks,
  696. 0 tests completed and failed residual checks,
  697. 0 tests skipped because of illegal input values.
  698. --------------------------------------------------------------------------------
  699.  
  700. End of Tests.
  701. ================================================================================
  702. ================================================================================
  703. HPLinpack 2.1 -- High-Performance Linpack benchmark -- October 26, 2012
  704. Written by A. Petitet and R. Clint Whaley, Innovative Computing Laboratory, UTK
  705. Modified by Piotr Luszczek, Innovative Computing Laboratory, UTK
  706. Modified by Julien Langou, University of Colorado Denver
  707. ================================================================================
  708.  
  709. An explanation of the input/output parameters follows:
  710. T/V : Wall time / encoded variant.
  711. N : The order of the coefficient matrix A.
  712. NB : The partitioning blocking factor.
  713. P : The number of process rows.
  714. Q : The number of process columns.
  715. Time : Time in seconds to solve the linear system.
  716. Gflops : Rate of execution for solving the linear system.
  717.  
  718. The following parameter values will be used:
  719.  
  720. N : 4096
  721. NB : 256
  722. PMAP : Row-major process mapping
  723. P : 1
  724. Q : 1
  725. PFACT : Left
  726. NBMIN : 2
  727. NDIV : 2
  728. RFACT : Right
  729. BCAST : 2ring
  730. DEPTH : 0
  731. SWAP : Mix (threshold = 64)
  732. L1 : transposed form
  733. U : transposed form
  734. EQUIL : yes
  735. ALIGN : 8 double precision words
  736.  
  737. --------------------------------------------------------------------------------
  738.  
  739. - The matrix A is randomly generated for each test.
  740. - The following scaled residual check will be computed:
  741. ||Ax-b||_oo / ( eps * ( || x ||_oo * || A ||_oo + || b ||_oo ) * N )
  742. - The relative machine precision (eps) is taken to be 1.110223e-16
  743. - Computational tests pass if scaled residuals are less than 0.0
  744.  
  745. ================================================================================
  746. T/V N NB P Q Time Gflops
  747. --------------------------------------------------------------------------------
  748. WR02R2L2 4096 256 1 1 30.78 1.489e+00
  749. HPL_pdgesv() start time Sun Aug 6 13:26:20 2017
  750.  
  751. HPL_pdgesv() end time Sun Aug 6 13:26:51 2017
  752.  
  753. --------------------------------------------------------------------------------
  754. ||Ax-b||_oo/(eps*(||A||_oo*||x||_oo+||b||_oo)*N)= 0.0048034 ...... PASSED
  755. ================================================================================
  756.  
  757. Finished 1 tests with the following results:
  758. 1 tests completed and passed residual checks,
  759. 0 tests completed and failed residual checks,
  760. 0 tests skipped because of illegal input values.
  761. --------------------------------------------------------------------------------
  762.  
  763. End of Tests.
  764. ================================================================================
  765. ================================================================================
  766. HPLinpack 2.1 -- High-Performance Linpack benchmark -- October 26, 2012
  767. Written by A. Petitet and R. Clint Whaley, Innovative Computing Laboratory, UTK
  768. Modified by Piotr Luszczek, Innovative Computing Laboratory, UTK
  769. Modified by Julien Langou, University of Colorado Denver
  770. ================================================================================
  771.  
  772. An explanation of the input/output parameters follows:
  773. T/V : Wall time / encoded variant.
  774. N : The order of the coefficient matrix A.
  775. NB : The partitioning blocking factor.
  776. P : The number of process rows.
  777. Q : The number of process columns.
  778. Time : Time in seconds to solve the linear system.
  779. Gflops : Rate of execution for solving the linear system.
  780.  
  781. The following parameter values will be used:
  782.  
  783. N : 4096
  784. NB : 256
  785. PMAP : Row-major process mapping
  786. P : 1
  787. Q : 1
  788. PFACT : Left
  789. NBMIN : 2
  790. NDIV : 2
  791. RFACT : Right
  792. BCAST : 2ring
  793. DEPTH : 0
  794. SWAP : Mix (threshold = 64)
  795. L1 : transposed form
  796. U : transposed form
  797. EQUIL : yes
  798. ALIGN : 8 double precision words
  799.  
  800. --------------------------------------------------------------------------------
  801.  
  802. - The matrix A is randomly generated for each test.
  803. - The following scaled residual check will be computed:
  804. ||Ax-b||_oo / ( eps * ( || x ||_oo * || A ||_oo + || b ||_oo ) * N )
  805. - The relative machine precision (eps) is taken to be 1.110223e-16
  806. - Computational tests pass if scaled residuals are less than 0.0
  807.  
  808. ================================================================================
  809. T/V N NB P Q Time Gflops
  810. --------------------------------------------------------------------------------
  811. WR02R2L2 4096 256 1 1 30.95 1.481e+00
  812. HPL_pdgesv() start time Sun Aug 6 13:26:56 2017
  813.  
  814. HPL_pdgesv() end time Sun Aug 6 13:27:27 2017
  815.  
  816. --------------------------------------------------------------------------------
  817. ||Ax-b||_oo/(eps*(||A||_oo*||x||_oo+||b||_oo)*N)= 0.0048034 ...... PASSED
  818. ================================================================================
  819.  
  820. Finished 1 tests with the following results:
  821. 1 tests completed and passed residual checks,
  822. 0 tests completed and failed residual checks,
  823. 0 tests skipped because of illegal input values.
  824. --------------------------------------------------------------------------------
  825.  
  826. End of Tests.
  827. ================================================================================
  828. ================================================================================
  829. HPLinpack 2.1 -- High-Performance Linpack benchmark -- October 26, 2012
  830. Written by A. Petitet and R. Clint Whaley, Innovative Computing Laboratory, UTK
  831. Modified by Piotr Luszczek, Innovative Computing Laboratory, UTK
  832. Modified by Julien Langou, University of Colorado Denver
  833. ================================================================================
  834.  
  835. An explanation of the input/output parameters follows:
  836. T/V : Wall time / encoded variant.
  837. N : The order of the coefficient matrix A.
  838. NB : The partitioning blocking factor.
  839. P : The number of process rows.
  840. Q : The number of process columns.
  841. Time : Time in seconds to solve the linear system.
  842. Gflops : Rate of execution for solving the linear system.
  843.  
  844. The following parameter values will be used:
  845.  
  846. N : 4096
  847. NB : 256
  848. PMAP : Row-major process mapping
  849. P : 1
  850. Q : 1
  851. PFACT : Left
  852. NBMIN : 2
  853. NDIV : 2
  854. RFACT : Right
  855. BCAST : 2ring
  856. DEPTH : 0
  857. SWAP : Mix (threshold = 64)
  858. L1 : transposed form
  859. U : transposed form
  860. EQUIL : yes
  861. ALIGN : 8 double precision words
  862.  
  863. --------------------------------------------------------------------------------
  864.  
  865. - The matrix A is randomly generated for each test.
  866. - The following scaled residual check will be computed:
  867. ||Ax-b||_oo / ( eps * ( || x ||_oo * || A ||_oo + || b ||_oo ) * N )
  868. - The relative machine precision (eps) is taken to be 1.110223e-16
  869. - Computational tests pass if scaled residuals are less than 0.0
  870.  
  871. ================================================================================
  872. T/V N NB P Q Time Gflops
  873. --------------------------------------------------------------------------------
  874. WR02R2L2 4096 256 1 1 30.89 1.484e+00
  875. HPL_pdgesv() start time Sun Aug 6 13:27:32 2017
  876.  
  877. HPL_pdgesv() end time Sun Aug 6 13:28:03 2017
  878.  
  879. --------------------------------------------------------------------------------
  880. ||Ax-b||_oo/(eps*(||A||_oo*||x||_oo+||b||_oo)*N)= 0.0048034 ...... PASSED
  881. ================================================================================
  882.  
  883. Finished 1 tests with the following results:
  884. 1 tests completed and passed residual checks,
  885. 0 tests completed and failed residual checks,
  886. 0 tests skipped because of illegal input values.
  887. --------------------------------------------------------------------------------
  888.  
  889. End of Tests.
  890. ================================================================================
  891. ================================================================================
  892. HPLinpack 2.1 -- High-Performance Linpack benchmark -- October 26, 2012
  893. Written by A. Petitet and R. Clint Whaley, Innovative Computing Laboratory, UTK
  894. Modified by Piotr Luszczek, Innovative Computing Laboratory, UTK
  895. Modified by Julien Langou, University of Colorado Denver
  896. ================================================================================
  897.  
  898. An explanation of the input/output parameters follows:
  899. T/V : Wall time / encoded variant.
  900. N : The order of the coefficient matrix A.
  901. NB : The partitioning blocking factor.
  902. P : The number of process rows.
  903. Q : The number of process columns.
  904. Time : Time in seconds to solve the linear system.
  905. Gflops : Rate of execution for solving the linear system.
  906.  
  907. The following parameter values will be used:
  908.  
  909. N : 4096
  910. NB : 256
  911. PMAP : Row-major process mapping
  912. P : 1
  913. Q : 1
  914. PFACT : Left
  915. NBMIN : 2
  916. NDIV : 2
  917. RFACT : Right
  918. BCAST : 2ring
  919. DEPTH : 0
  920. SWAP : Mix (threshold = 64)
  921. L1 : transposed form
  922. U : transposed form
  923. EQUIL : yes
  924. ALIGN : 8 double precision words
  925.  
  926. --------------------------------------------------------------------------------
  927.  
  928. - The matrix A is randomly generated for each test.
  929. - The following scaled residual check will be computed:
  930. ||Ax-b||_oo / ( eps * ( || x ||_oo * || A ||_oo + || b ||_oo ) * N )
  931. - The relative machine precision (eps) is taken to be 1.110223e-16
  932. - Computational tests pass if scaled residuals are less than 0.0
  933.  
  934. ================================================================================
  935. T/V N NB P Q Time Gflops
  936. --------------------------------------------------------------------------------
  937. WR02R2L2 4096 256 1 1 30.79 1.489e+00
  938. HPL_pdgesv() start time Sun Aug 6 13:28:08 2017
  939.  
  940. HPL_pdgesv() end time Sun Aug 6 13:28:39 2017
  941.  
  942. --------------------------------------------------------------------------------
  943. ||Ax-b||_oo/(eps*(||A||_oo*||x||_oo+||b||_oo)*N)= 0.0048034 ...... PASSED
  944. ================================================================================
  945.  
  946. Finished 1 tests with the following results:
  947. 1 tests completed and passed residual checks,
  948. 0 tests completed and failed residual checks,
  949. 0 tests skipped because of illegal input values.
  950. --------------------------------------------------------------------------------
  951.  
  952. End of Tests.
  953. ================================================================================
Advertisement
Add Comment
Please, Sign In to add comment
Advertisement
Public Pastes
Advertisement
We use cookies for various purposes including analytics. By continuing to use Pastebin, you agree to our use of cookies as described in the Cookies Policy.  OK, I Understand
Not a member of Pastebin yet?
Sign Up, it unlocks many cool features!