b_eff = 9493.817 MB/s = 632.921 * 15 PEs with 256 MB/PE on SUPER-UX hwwsx4 9.1 Rev1 SX-4 SECTION-HEAD-BEGIN b_eff.c, Revision 3.2 from Nov. 07, 1999 MEMORY_PER_PROCESSOR = 256 MBytes [1M = 1024*1024] 1-dim-paterns: size = 15 1-dim-paterns: size = 15 2-dim-paterns: size = 5 * 3 3-dim-paterns: size = 3 * 2 * 2 Used message lengths (and used loop length for each message lengths): msglng= 1 (300), 2 (300), 4 (300), 8 (300), 16 (300), 32 (300), 64 (300), 128 (300), 256 (300), 512 (300), 1024 (300), 2048 (300), 4096 (300), 8933 (234), 19484 (107), 42495 (49), 92682 (22), 202141 (10), 440872 (4), 961548 (2), 2097152 (1), SECTION-HEAD-END SECTION-LOOP-BEGIN measurment loop: elapsed time = 95.129 sec sum of max elapsed time per entries above = 94.356 sec difference = 0.772 sec = 0.8% The difference is less than 5 % SECTION-LOOP-END SECTION-PATTERN-BEGIN Pattern parameters: p00 ring-7*2&+1 => 1 sendrecv_calls with 15 messages, i.e. 4.2 msgs/used node, all nodes are used p01 ring-4*4&-1 => 2 sendrecv_calls with 30 messages, i.e. 4.2 msgs/used node, all nodes are used p02 ring-2*8&-1 => 2 sendrecv_calls with 30 messages, i.e. 4.2 msgs/used node, all nodes are used p03 ring-1*15fix => 2 sendrecv_calls with 30 messages, i.e. 4.2 msgs/used node, all nodes are used p04 ring-1*15fix => 2 sendrecv_calls with 30 messages, i.e. 4.2 msgs/used node, all nodes are used p05 ring-1*15fix => 2 sendrecv_calls with 30 messages, i.e. 4.2 msgs/used node, all nodes are used p06 random-cyc-1dim => 2 sendrecv_calls with 30 messages, i.e. 4.2 msgs/used node, all nodes are used p07 random-cyc-1dim => 2 sendrecv_calls with 30 messages, i.e. 4.2 msgs/used node, all nodes are used p08 random-cyc-1dim => 2 sendrecv_calls with 30 messages, i.e. 4.2 msgs/used node, all nodes are used p09 random-cyc-1dim => 2 sendrecv_calls with 30 messages, i.e. 4.2 msgs/used node, all nodes are used p10 random-cyc-1dim => 2 sendrecv_calls with 30 messages, i.e. 4.2 msgs/used node, all nodes are used p11 random-cyc-1dim => 2 sendrecv_calls with 30 messages, i.e. 4.2 msgs/used node, all nodes are used p12 random-cyc-1dim => 2 sendrecv_calls with 30 messages, i.e. 4.2 msgs/used node, all nodes are used p13 random-cyc-1dim => 2 sendrecv_calls with 30 messages, i.e. 4.2 msgs/used node, all nodes are used p14 random-cyc-1dim => 2 sendrecv_calls with 30 messages, i.e. 4.2 msgs/used node, all nodes are used p15 random-cyc-1dim => 2 sendrecv_calls with 30 messages, i.e. 4.2 msgs/used node, all nodes are used p16 worst-cyc-1dim => 2 sendrecv_calls with 30 messages, i.e. 4.2 msgs/used node, all nodes are used p17 best bi-section => 2 sendrecv_calls with 14 messages, i.e. 4.2 msgs/used node, 1 nodes are UNUSED p18 worst bi-section => 2 sendrecv_calls with 14 messages, i.e. 4.2 msgs/used node, 1 nodes are UNUSED p19 acyclic-1dim-all => 2 sendrecv_calls with 28 messages, i.e. 4.2 msgs/used node, all nodes are used p20 acyclic-2dim-all => 4 sendrecv_calls with 44 messages, i.e. 4.2 msgs/used node, all nodes are used p21 acyclic-3dim-all => 6 sendrecv_calls with 40 messages, i.e. 4.2 msgs/used node, 3 nodes are UNUSED p22 cyclic-1dim-all => 2 sendrecv_calls with 30 messages, i.e. 4.2 msgs/used node, all nodes are used p23 cyclic-2dim-all => 4 sendrecv_calls with 60 messages, i.e. 4.2 msgs/used node, all nodes are used p24 cyclic-3dim-all => 4 sendrecv_calls with 48 messages, i.e. 4.2 msgs/used node, 3 nodes are UNUSED SECTION-PATTERN-END Printing accumulated bandwidth / number of processes = bandwidth per process (in MBytes/sec = 1e6 Bytes/sec) Additionally the accumulated value processes is printed in the last column Only the cyclic and random patterns are used to compute b_eff (except p00), the other patterns are only as information. SECTION-BY-METHODS-BEGIN 2nd analysis path, only as information, last row: for all mthd: logavg_pat ( avg_L ( max_rep ( b(L) ) ) ) || logavg_pat ( max_mthd ( avg_L ( max_rep ( b(L) ) ) ) ) pattern / methods: Sndrcv Alltoal non-blk -> best method -> accumulated p00 ring-7*2&+1 : 640.327 164.420 523.816 -> 640.327 -> 9604.907 MByte/s p01 ring-4*4&-1 : 641.763 298.621 520.681 -> 641.763 -> 9626.438 MByte/s p02 ring-2*8&-1 : 637.137 299.092 535.945 -> 637.137 -> 9557.053 MByte/s p03 ring-1*15fix : 631.146 448.563 546.169 -> 631.146 -> 9467.183 MByte/s p04 ring-1*15fix : 620.787 447.837 544.240 -> 620.787 -> 9311.805 MByte/s p05 ring-1*15fix : 625.275 448.093 545.030 -> 625.275 -> 9379.118 MByte/s p06 random-cyc-1dim : 606.632 449.577 538.495 -> 606.632 -> 9099.481 MByte/s p07 random-cyc-1dim : 624.453 448.987 521.508 -> 624.453 -> 9366.797 MByte/s p08 random-cyc-1dim : 627.223 448.789 504.169 -> 627.223 -> 9408.339 MByte/s p09 random-cyc-1dim : 641.561 449.745 552.573 -> 641.561 -> 9623.409 MByte/s p10 random-cyc-1dim : 623.029 448.638 515.313 -> 623.029 -> 9345.428 MByte/s p11 random-cyc-1dim : 610.528 448.908 514.355 -> 610.528 -> 9157.916 MByte/s p12 random-cyc-1dim : 639.646 449.422 500.928 -> 639.646 -> 9594.697 MByte/s p13 random-cyc-1dim : 635.406 449.709 514.253 -> 635.406 -> 9531.089 MByte/s p14 random-cyc-1dim : 643.048 449.873 508.023 -> 643.048 -> 9645.727 MByte/s p15 random-cyc-1dim : 642.815 449.625 520.453 -> 642.815 -> 9642.222 MByte/s -- additional patterns that are not used to compute b_eff: p16 worst-cyc-1dim : 629.474 449.709 525.146 -> 629.474 -> 9442.105 MByte/s p17 best bi-section : 375.406 256.244 522.542 -> 522.542 -> 7838.135 MByte/s p18 worst bi-section : 375.167 180.166 522.486 -> 522.486 -> 7837.294 MByte/s p19 acyclic-1dim-all : 587.739 423.237 519.244 -> 587.739 -> 8816.081 MByte/s p20 acyclic-2dim-all : 454.485 421.105 401.546 -> 454.485 -> 6817.269 MByte/s p21 acyclic-3dim-all : 328.691 316.114 390.463 -> 390.463 -> 5856.952 MByte/s p22 cyclic-1dim-all : 635.239 451.725 551.750 -> 635.239 -> 9528.579 MByte/s p23 cyclic-2dim-all : 624.756 453.112 490.561 -> 624.756 -> 9371.340 MByte/s p24 cyclic-3dim-all : 515.596 377.570 396.806 -> 515.596 -> 7733.935 MByte/s log_avg of all rings : 632.691 331.282 535.881 || 632.691 -> 9490.372 MByte/s log_avg of all random : 629.306 449.327 518.795 || 629.306 -> 9439.590 MByte/s log_avg(ring,random) : 630.996 385.816 527.269 ||(630.996 -> 9464.947)MByte/s SECTION-BY-METHODS-END SECTION-BY-REPETITIONS-BEGIN 3rd analysis path, only as information, last row: for all rep.:logavg_pat ( avg_L ( max_mthd ( b(L) ) ) ) || logavg_pat ( max_rep ( avg_L ( max_mthd ( b(L) ) ) ) ) pattern / repetition: rep.0 rep.1 rep.2 -> best repetition -> accumulated p00 ring-7*2&+1 : 569.315 605.714 637.181 -> 637.181 -> 9557.720 MByte/s p01 ring-4*4&-1 : 554.977 585.801 612.933 -> 612.933 -> 9193.989 MByte/s p02 ring-2*8&-1 : 550.103 587.400 608.730 -> 608.730 -> 9130.943 MByte/s p03 ring-1*15fix : 584.987 624.368 622.620 -> 624.368 -> 9365.521 MByte/s p04 ring-1*15fix : 559.410 599.639 616.643 -> 616.643 -> 9249.652 MByte/s p05 ring-1*15fix : 567.524 596.927 615.890 -> 615.890 -> 9238.352 MByte/s p06 random-cyc-1dim : 550.274 612.109 600.859 -> 612.109 -> 9181.628 MByte/s p07 random-cyc-1dim : 559.471 595.648 621.995 -> 621.995 -> 9329.924 MByte/s p08 random-cyc-1dim : 582.652 608.057 626.390 -> 626.390 -> 9395.843 MByte/s p09 random-cyc-1dim : 566.183 605.766 613.185 -> 613.185 -> 9197.781 MByte/s p10 random-cyc-1dim : 557.893 615.597 587.769 -> 615.597 -> 9233.956 MByte/s p11 random-cyc-1dim : 552.355 614.579 578.068 -> 614.579 -> 9218.681 MByte/s p12 random-cyc-1dim : 547.110 638.721 571.283 -> 638.721 -> 9580.808 MByte/s p13 random-cyc-1dim : 558.122 631.226 596.078 -> 631.226 -> 9468.396 MByte/s p14 random-cyc-1dim : 615.375 611.943 614.838 -> 615.375 -> 9230.630 MByte/s p15 random-cyc-1dim : 548.885 613.595 609.189 -> 613.595 -> 9203.926 MByte/s -- additional patterns that are not used to compute b_eff: p16 worst-cyc-1dim : 568.683 597.870 593.798 -> 597.870 -> 8968.043 MByte/s p17 best bi-section : 516.668 520.595 519.412 -> 520.595 -> 7808.923 MByte/s p18 worst bi-section : 513.952 523.635 519.250 -> 523.635 -> 7854.521 MByte/s p19 acyclic-1dim-all : 553.393 591.682 581.399 -> 591.682 -> 8875.228 MByte/s p20 acyclic-2dim-all : 445.420 464.514 447.497 -> 464.514 -> 6967.711 MByte/s p21 acyclic-3dim-all : 361.051 383.987 373.802 -> 383.987 -> 5759.804 MByte/s p22 cyclic-1dim-all : 559.873 633.991 598.986 -> 633.991 -> 9509.863 MByte/s p23 cyclic-2dim-all : 516.761 612.447 595.128 -> 612.447 -> 9186.700 MByte/s p24 cyclic-3dim-all : 440.352 492.048 503.116 -> 503.116 -> 7546.745 MByte/s log_avg of all rings : 564.272 599.838 618.933 || 619.222 -> 9288.330 MByte/s log_avg of all random : 563.497 614.615 601.707 || 620.218 -> 9303.270 MByte/s log_avg(ring,random) : 563.884 607.181 610.259 ||(619.720 -> 9295.797)MByte/s SECTION-BY-REPETITIONS-END SECTION-BY-MTHD-MSGLNG-BEGIN 4th analysis path, only as information, last 3 rows: for all methods and some msglng L: logavg_pat ( max_rep ( b(L) ) ) for all methods: logavg_pat ( avg_L ( max_rep ( b(L) ) ) ) pattern & method / msg: 1 16 256 4096 92682 2097152 -> average -> accumulated p00 ring-7*2&+1 p00 method 0 : 0.049 0.808 12.733 157.675 1435.223 3514.937 -> 640.327 -> 9604.907 MByte/s p00 method 1 : 0.006 0.116 1.836 28.749 376.041 841.319 -> 164.420 -> 2466.295 MByte/s p00 method 2 : 0.034 0.531 8.645 119.395 1207.231 2867.673 -> 523.816 -> 7857.244 MByte/s p01 ring-4*4&-1 p01 method 0 : 0.049 0.812 12.708 156.194 1451.319 3524.081 -> 641.763 -> 9626.438 MByte/s p01 method 1 : 0.012 0.230 3.613 56.832 723.467 1408.774 -> 298.621 -> 4479.311 MByte/s p01 method 2 : 0.039 0.609 9.799 131.297 1359.369 2397.293 -> 520.681 -> 7810.218 MByte/s p02 ring-2*8&-1 p02 method 0 : 0.049 0.812 12.785 152.899 1441.143 3522.872 -> 637.137 -> 9557.053 MByte/s p02 method 1 : 0.012 0.230 3.608 57.430 725.550 1401.746 -> 299.092 -> 4486.373 MByte/s p02 method 2 : 0.038 0.601 9.721 133.273 1344.975 2816.860 -> 535.945 -> 8039.180 MByte/s p03 ring-1*15fix p03 method 0 : 0.049 0.811 12.810 153.120 1437.262 3541.457 -> 631.146 -> 9467.183 MByte/s p03 method 1 : 0.012 0.230 3.537 58.086 731.883 3006.497 -> 448.563 -> 6728.447 MByte/s p03 method 2 : 0.039 0.601 9.718 130.060 1347.168 2746.631 -> 546.169 -> 8192.536 MByte/s p04 ring-1*15fix p04 method 0 : 0.049 0.811 12.781 157.326 1440.133 3533.510 -> 620.787 -> 9311.805 MByte/s p04 method 1 : 0.012 0.230 3.613 58.016 726.722 3001.867 -> 447.837 -> 6717.560 MByte/s p04 method 2 : 0.038 0.600 9.745 127.482 1351.118 2747.582 -> 544.240 -> 8163.603 MByte/s p05 ring-1*15fix p05 method 0 : 0.049 0.811 12.796 157.518 1442.762 3507.388 -> 625.275 -> 9379.118 MByte/s p05 method 1 : 0.012 0.227 3.588 57.864 727.386 2977.962 -> 448.093 -> 6721.397 MByte/s p05 method 2 : 0.036 0.596 9.828 129.473 1348.419 2774.524 -> 545.030 -> 8175.448 MByte/s p06 random-cyc-1dim p06 method 0 : 0.049 0.811 12.603 159.222 1444.684 2763.003 -> 606.632 -> 9099.481 MByte/s p06 method 1 : 0.012 0.227 3.596 57.973 725.582 3008.896 -> 449.577 -> 6743.650 MByte/s p06 method 2 : 0.036 0.623 9.718 133.353 1354.420 2596.526 -> 538.495 -> 8077.429 MByte/s p07 random-cyc-1dim p07 method 0 : 0.049 0.812 12.817 158.643 1434.500 3538.207 -> 624.453 -> 9366.797 MByte/s p07 method 1 : 0.012 0.228 3.648 58.002 724.360 2997.576 -> 448.987 -> 6734.809 MByte/s p07 method 2 : 0.036 0.604 9.643 124.914 1344.011 2578.356 -> 521.508 -> 7822.625 MByte/s p08 random-cyc-1dim p08 method 0 : 0.049 0.810 12.774 159.374 1445.282 3544.762 -> 627.223 -> 9408.339 MByte/s p08 method 1 : 0.012 0.228 3.526 57.441 724.850 3021.014 -> 448.789 -> 6731.830 MByte/s p08 method 2 : 0.037 0.606 9.838 130.259 1356.273 2229.874 -> 504.169 -> 7562.534 MByte/s p09 random-cyc-1dim p09 method 0 : 0.048 0.801 12.576 159.107 1436.424 3540.668 -> 641.561 -> 9623.409 MByte/s p09 method 1 : 0.012 0.230 3.495 58.025 729.552 3023.680 -> 449.745 -> 6746.179 MByte/s p09 method 2 : 0.038 0.599 9.702 129.545 1355.825 2869.164 -> 552.573 -> 8288.596 MByte/s p10 random-cyc-1dim p10 method 0 : 0.049 0.813 12.558 158.934 1438.439 3142.780 -> 623.029 -> 9345.428 MByte/s p10 method 1 : 0.012 0.230 3.515 57.971 725.651 3027.644 -> 448.638 -> 6729.563 MByte/s p10 method 2 : 0.036 0.607 9.517 130.545 1352.372 2348.687 -> 515.313 -> 7729.690 MByte/s p11 random-cyc-1dim p11 method 0 : 0.049 0.812 12.809 158.778 1437.250 2863.882 -> 610.528 -> 9157.916 MByte/s p11 method 1 : 0.012 0.230 3.584 57.514 728.625 3007.911 -> 448.908 -> 6733.619 MByte/s p11 method 2 : 0.039 0.607 9.667 132.028 1353.399 2398.939 -> 514.355 -> 7715.330 MByte/s p12 random-cyc-1dim p12 method 0 : 0.049 0.811 12.157 156.849 1432.525 3512.959 -> 639.646 -> 9594.697 MByte/s p12 method 1 : 0.011 0.230 3.584 57.852 723.524 3022.408 -> 449.422 -> 6741.333 MByte/s p12 method 2 : 0.039 0.650 9.861 133.509 1351.705 2150.855 -> 500.928 -> 7513.924 MByte/s p13 random-cyc-1dim p13 method 0 : 0.048 0.813 12.821 159.248 1437.550 3531.534 -> 635.406 -> 9531.089 MByte/s p13 method 1 : 0.012 0.230 3.480 58.022 727.152 3020.232 -> 449.709 -> 6745.639 MByte/s p13 method 2 : 0.039 0.605 10.182 130.224 1352.333 2413.682 -> 514.253 -> 7713.801 MByte/s p14 random-cyc-1dim p14 method 0 : 0.049 0.814 12.829 158.945 1433.987 3530.702 -> 643.048 -> 9645.727 MByte/s p14 method 1 : 0.012 0.230 3.598 58.021 727.349 3008.136 -> 449.873 -> 6748.091 MByte/s p14 method 2 : 0.039 0.626 9.708 131.038 1352.587 2218.354 -> 508.023 -> 7620.345 MByte/s p15 random-cyc-1dim p15 method 0 : 0.044 0.812 12.793 158.898 1435.522 3527.044 -> 642.815 -> 9642.222 MByte/s p15 method 1 : 0.012 0.228 3.603 58.060 723.464 3026.613 -> 449.625 -> 6744.374 MByte/s p15 method 2 : 0.039 0.628 9.551 134.449 1353.586 2625.812 -> 520.453 -> 7806.792 MByte/s -- additional patterns that are not used to compute b_eff: p16 worst-cyc-1dim p16 method 0 : 0.048 0.812 12.782 157.202 1439.182 3521.737 -> 629.474 -> 9442.105 MByte/s p16 method 1 : 0.011 0.229 3.578 57.997 728.738 3016.582 -> 449.709 -> 6745.629 MByte/s p16 method 2 : 0.039 0.642 9.756 131.259 1352.448 2571.211 -> 525.146 -> 7877.192 MByte/s p17 best bi-section p17 method 0 : 0.035 0.582 8.913 105.825 780.756 2243.793 -> 375.406 -> 5631.089 MByte/s p17 method 1 : 0.006 0.108 1.664 27.345 410.630 1405.668 -> 256.244 -> 3843.665 MByte/s p17 method 2 : 0.025 0.393 6.167 88.936 1009.689 3165.704 -> 522.542 -> 7838.135 MByte/s p18 worst bi-section p18 method 0 : 0.035 0.580 9.086 106.064 780.730 2239.686 -> 375.167 -> 5627.503 MByte/s p18 method 1 : 0.006 0.108 1.700 27.384 414.674 395.803 -> 180.166 -> 2702.488 MByte/s p18 method 2 : 0.025 0.392 6.110 89.988 1015.512 3162.348 -> 522.486 -> 7837.294 MByte/s p19 acyclic-1dim-all p19 method 0 : 0.045 0.757 11.945 148.781 1353.685 3293.946 -> 587.739 -> 8816.081 MByte/s p19 method 1 : 0.011 0.214 3.411 54.077 689.086 2820.848 -> 423.237 -> 6348.557 MByte/s p19 method 2 : 0.033 0.560 9.167 125.275 1258.631 2591.421 -> 519.244 -> 7788.662 MByte/s p20 acyclic-2dim-all p20 method 0 : 0.035 0.598 9.439 117.189 1072.574 2353.749 -> 454.485 -> 6817.269 MByte/s p20 method 1 : 0.016 0.332 5.193 83.239 826.945 2447.065 -> 421.105 -> 6316.575 MByte/s p20 method 2 : 0.029 0.487 7.659 102.246 1009.402 1922.053 -> 401.546 -> 6023.192 MByte/s p21 acyclic-3dim-all p21 method 0 : 0.027 0.451 7.086 85.935 705.760 1934.376 -> 328.691 -> 4930.361 MByte/s p21 method 1 : 0.016 0.334 5.240 82.898 777.665 1565.293 -> 316.114 -> 4741.707 MByte/s p21 method 2 : 0.022 0.374 5.817 81.015 874.501 2189.844 -> 390.463 -> 5856.952 MByte/s p22 cyclic-1dim-all p22 method 0 : 0.048 0.799 12.796 156.357 1446.627 3511.947 -> 635.239 -> 9528.579 MByte/s p22 method 1 : 0.011 0.230 3.607 58.141 730.129 3013.132 -> 451.725 -> 6775.872 MByte/s p22 method 2 : 0.036 0.605 9.770 131.075 1351.458 2758.365 -> 551.750 -> 8276.255 MByte/s p23 cyclic-2dim-all p23 method 0 : 0.049 0.808 12.849 159.394 1449.394 3516.092 -> 624.756 -> 9371.340 MByte/s p23 method 1 : 0.023 0.430 7.096 112.951 1086.649 2312.838 -> 453.112 -> 6796.674 MByte/s p23 method 2 : 0.033 0.581 8.879 121.486 1306.777 2170.489 -> 490.561 -> 7358.418 MByte/s p24 cyclic-3dim-all p24 method 0 : 0.040 0.641 10.109 128.305 1169.774 2808.203 -> 515.596 -> 7733.935 MByte/s p24 method 1 : 0.019 0.396 6.261 99.734 912.182 1876.487 -> 377.570 -> 5663.549 MByte/s p24 method 2 : 0.025 0.420 7.410 95.406 1059.685 1839.030 -> 396.806 -> 5952.094 MByte/s log_avg of all rings - ring, method 0 : 0.049 0.811 12.769 155.775 1441.298 3524.023 || 632.691 -> 9490.372 MByte/s - ring, method 1 : 0.011 0.205 3.211 51.333 651.353 1883.535 || 331.282 -> 4969.236 MByte/s - ring, method 2 : 0.037 0.589 9.566 128.418 1325.247 2720.566 || 535.881 -> 8038.213 MByte/s log_avg of all random - random, method 0 : 0.048 0.811 12.672 158.798 1437.611 3335.811 || 629.306 -> 9439.590 MByte/s - random, method 1 : 0.012 0.229 3.562 57.888 726.008 3016.396 || 449.327 -> 6739.906 MByte/s - random, method 2 : 0.038 0.615 9.737 130.961 1352.647 2433.997 || 518.795 -> 7781.930 MByte/s log_avg(ring,random) - average, method 0 : 0.049 0.811 12.720 157.280 1439.453 3428.626 || 630.996 -> 9464.947 MByte/s - average, method 1 : 0.011 0.217 3.382 54.512 687.668 2383.587 || 385.816 -> 5787.243 MByte/s - average, method 2 : 0.038 0.602 9.651 129.683 1338.877 2573.296 || 527.269 -> 7909.033 MByte/s SECTION-BY-MTHD-MSGLNG-END SECTION-BY-MSGLNG-BEGIN 5th analysis path only as information: logavg_pat ( max_mthd ( max_rep ( b(L) ) ) ) and for all methods: logavg_pat ( max_rep ( b(L) ) ) msg length|accumulated| effective bandwidth per process: msg length| effective| average rings random methd_1 methd_2 methd_3 | bandwidth| crt,rnd only only Sendrcv Alltoall non-blk 1 0.730 0.049 0.049 0.048 0.049 0.011 0.038 2 1.445 0.096 0.096 0.097 0.096 0.023 0.073 4 2.916 0.194 0.193 0.195 0.194 0.053 0.149 8 6.004 0.400 0.402 0.399 0.400 0.105 0.296 16 12.162 0.811 0.811 0.811 0.811 0.217 0.602 32 23.977 1.598 1.599 1.597 1.598 0.429 1.194 64 48.053 3.204 3.204 3.203 3.204 0.850 2.432 128 95.285 6.352 6.352 6.353 6.352 1.669 4.780 256 190.806 12.720 12.769 12.672 12.720 3.382 9.651 512 378.374 25.225 25.222 25.227 25.225 6.758 18.844 1024 759.287 50.619 50.732 50.506 50.619 13.514 37.961 2048 1183.573 78.905 78.560 79.251 78.905 27.047 65.336 4096 2359.194 157.280 155.775 158.798 157.280 54.512 129.683 8933 3134.574 208.972 209.021 208.922 208.972 98.226 185.845 19484 8115.622 541.041 540.527 541.556 541.041 197.103 481.091 42495 12390.277 826.018 827.058 824.980 826.018 365.434 755.077 92682 21591.797 1439.453 1441.298 1437.611 1439.453 687.668 1338.877 202141 24886.626 1659.108 1657.055 1661.164 1659.108 1152.716 1484.788 440872 41618.892 2774.593 2772.379 2776.808 2774.593 1531.686 2298.393 961548 30677.454 2045.164 2004.362 2086.796 2022.570 1506.873 1673.039 2097152 51775.957 3451.730 3524.023 3380.921 3428.626 2383.587 2573.296 SECTION-BY-MSGLNG-END SECTION-BY-PATTERN-MSGLNG-BEGIN official (1st) analysis path -- see last column: (only columns of some message lengths are printed) logavg_pat ( avg_L ( max_mthd ( max_rep ( b(L) ) ) ) ) pattern / msg-length: 1 16 256 4096 92682 2097152 -> average -> accumulated p00 ring-7*2&+1 : 0.049 0.808 12.733 157.675 1435.223 3514.937 -> 640.327 -> 9604.907 MByte/s p01 ring-4*4&-1 : 0.049 0.812 12.708 156.194 1451.319 3524.081 -> 641.763 -> 9626.438 MByte/s p02 ring-2*8&-1 : 0.049 0.812 12.785 152.899 1441.143 3522.872 -> 637.137 -> 9557.053 MByte/s p03 ring-1*15fix : 0.049 0.811 12.810 153.120 1437.262 3541.457 -> 631.146 -> 9467.183 MByte/s p04 ring-1*15fix : 0.049 0.811 12.781 157.326 1440.133 3533.510 -> 627.303 -> 9409.545 MByte/s p05 ring-1*15fix : 0.049 0.811 12.796 157.518 1442.762 3507.388 -> 626.768 -> 9401.515 MByte/s p06 random-cyc-1dim : 0.049 0.811 12.603 159.222 1444.684 3008.896 -> 618.341 -> 9275.119 MByte/s p07 random-cyc-1dim : 0.049 0.812 12.817 158.643 1434.500 3538.207 -> 628.109 -> 9421.635 MByte/s p08 random-cyc-1dim : 0.049 0.810 12.774 159.374 1445.282 3544.762 -> 629.363 -> 9440.452 MByte/s p09 random-cyc-1dim : 0.048 0.801 12.576 159.107 1436.424 3540.668 -> 641.561 -> 9623.409 MByte/s p10 random-cyc-1dim : 0.049 0.813 12.558 158.934 1438.439 3142.780 -> 623.029 -> 9345.428 MByte/s p11 random-cyc-1dim : 0.049 0.812 12.809 158.778 1437.250 3007.911 -> 617.386 -> 9260.794 MByte/s p12 random-cyc-1dim : 0.049 0.811 12.157 156.849 1432.525 3512.959 -> 639.646 -> 9594.697 MByte/s p13 random-cyc-1dim : 0.048 0.813 12.821 159.248 1437.550 3531.534 -> 635.406 -> 9531.089 MByte/s p14 random-cyc-1dim : 0.049 0.814 12.829 158.945 1433.987 3530.702 -> 643.048 -> 9645.727 MByte/s p15 random-cyc-1dim : 0.044 0.812 12.793 158.898 1435.522 3527.044 -> 642.815 -> 9642.222 MByte/s -- additional patterns that are not used to compute b_eff: p16 worst-cyc-1dim : 0.048 0.812 12.782 157.202 1439.182 3521.737 -> 629.474 -> 9442.105 MByte/s p17 best bi-section : 0.035 0.582 8.913 105.825 1009.689 3165.704 -> 525.247 -> 7878.705 MByte/s p18 worst bi-section : 0.035 0.580 9.086 106.064 1015.512 3162.348 -> 525.135 -> 7877.028 MByte/s p19 acyclic-1dim-all : 0.045 0.757 11.945 148.781 1353.685 3293.946 -> 594.762 -> 8921.423 MByte/s p20 acyclic-2dim-all : 0.035 0.598 9.439 117.189 1072.574 2447.065 -> 465.277 -> 6979.150 MByte/s p21 acyclic-3dim-all : 0.027 0.451 7.086 85.935 874.501 2189.844 -> 391.318 -> 5869.773 MByte/s p22 cyclic-1dim-all : 0.048 0.799 12.796 156.357 1446.627 3511.947 -> 635.239 -> 9528.579 MByte/s p23 cyclic-2dim-all : 0.049 0.808 12.849 159.394 1449.394 3516.092 -> 624.756 -> 9371.340 MByte/s p24 cyclic-3dim-all : 0.040 0.641 10.109 128.305 1169.774 2808.203 -> 515.596 -> 7733.935 MByte/s log_avg of all rings : 0.049 0.811 12.769 155.775 1441.298 3524.023 || 634.045 -> 9510.682 MByte/s log_avg of all random : 0.048 0.811 12.672 158.798 1437.611 3380.921 || 631.799 -> 9476.983 MByte/s log_avg(ring,random) : 0.049 0.811 12.720 157.280 1439.453 3451.730 || 632.921 -> 9493.817 MByte/s SECTION-BY-PATTERN-MSGLNG-END SECTION-BEFF-BEGIN The effective bandwidth is b_eff = 9493.817 MByte/s on 15 processes ( = 632.921 MByte/s * 15 processes) system parameters : 15 nodes, 256 MB/node system name: SUPER-UX hostname : hwwsx4 OS release : 9.1 OS version : Rev1 machine : SX-4 SECTION-BEFF-END b_eff = 9493.817 MB/s = 632.921 * 15 PEs with 256 MB/PE on SUPER-UX hwwsx4 9.1 Rev1 SX-4