b_eff = 1522.225 MB/s = 63.426 * 24 PEs with 128 MB/PE on sn6715 hwwt3e 2.0.4.71 unicosmk CRAY T3E SECTION-HEAD-BEGIN b_eff.c, Revision 3.2 from Nov. 07, 1999 MEMORY_PER_PROCESSOR = 128 MBytes [1M = 1024*1024] 1-dim-paterns: size = 24 1-dim-paterns: size = 24 2-dim-paterns: size = 6 * 4 3-dim-paterns: size = 4 * 3 * 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 (256), 8192 (128), 16384 (64), 32768 (32), 65536 (16), 131072 (8), 262144 (4), 524288 (2), 1048576 (1), SECTION-HEAD-END SECTION-LOOP-BEGIN measurment loop: elapsed time = 134.108 sec sum of max elapsed time per entries above = 136.299 sec difference = -2.191 sec = 1.6% The difference is less than 5 % SECTION-LOOP-END SECTION-PATTERN-BEGIN Pattern parameters: p00 ring-12*2fix => 1 sendrecv_calls with 24 messages, i.e. msgs/used node, all nodes are used p01 ring-6*4fix => 2 sendrecv_calls with 48 messages, i.e. msgs/used node, all nodes are used p02 ring-3*8fix => 2 sendrecv_calls with 48 messages, i.e. msgs/used node, all nodes are used p03 ring-1*24fix => 2 sendrecv_calls with 48 messages, i.e. msgs/used node, all nodes are used p04 ring-1*24fix => 2 sendrecv_calls with 48 messages, i.e. msgs/used node, all nodes are used p05 ring-1*24fix => 2 sendrecv_calls with 48 messages, i.e. msgs/used node, all nodes are used p06 random-cyc-1dim => 2 sendrecv_calls with 48 messages, i.e. msgs/used node, all nodes are used p07 random-cyc-1dim => 2 sendrecv_calls with 48 messages, i.e. msgs/used node, all nodes are used p08 random-cyc-1dim => 2 sendrecv_calls with 48 messages, i.e. msgs/used node, all nodes are used p09 random-cyc-1dim => 2 sendrecv_calls with 48 messages, i.e. msgs/used node, all nodes are used p10 random-cyc-1dim => 2 sendrecv_calls with 48 messages, i.e. msgs/used node, all nodes are used p11 random-cyc-1dim => 2 sendrecv_calls with 48 messages, i.e. msgs/used node, all nodes are used p12 random-cyc-1dim => 2 sendrecv_calls with 48 messages, i.e. msgs/used node, all nodes are used p13 random-cyc-1dim => 2 sendrecv_calls with 48 messages, i.e. msgs/used node, all nodes are used p14 random-cyc-1dim => 2 sendrecv_calls with 48 messages, i.e. msgs/used node, all nodes are used p15 random-cyc-1dim => 2 sendrecv_calls with 48 messages, i.e. msgs/used node, all nodes are used p16 worst-cyc-1dim => 2 sendrecv_calls with 48 messages, i.e. msgs/used node, all nodes are used p17 best bi-section => 2 sendrecv_calls with 24 messages, i.e. msgs/used node, all nodes are used p18 worst bi-section => 2 sendrecv_calls with 24 messages, i.e. msgs/used node, all nodes are used p19 acyclic-1dim-all => 2 sendrecv_calls with 46 messages, i.e. msgs/used node, all nodes are used p20 acyclic-2dim-all => 4 sendrecv_calls with 76 messages, i.e. msgs/used node, all nodes are used p21 acyclic-3dim-all => 6 sendrecv_calls with 92 messages, i.e. msgs/used node, all nodes are used p22 cyclic-1dim-all => 2 sendrecv_calls with 48 messages, i.e. msgs/used node, all nodes are used p23 cyclic-2dim-all => 4 sendrecv_calls with 96 messages, i.e. msgs/used node, all nodes are used p24 cyclic-3dim-all => 5 sendrecv_calls with 120 messages, i.e. msgs/used node, all nodes are used 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-12*2fix : 72.290 80.193 61.217 -> 80.193 -> 1924.632 MByte/s p01 ring-6*4fix : 61.416 82.282 46.577 -> 82.282 -> 1974.771 MByte/s p02 ring-3*8fix : 53.921 73.696 45.826 -> 73.696 -> 1768.702 MByte/s p03 ring-1*24fix : 56.418 85.382 47.606 -> 85.382 -> 2049.157 MByte/s p04 ring-1*24fix : 56.286 85.239 47.351 -> 85.239 -> 2045.745 MByte/s p05 ring-1*24fix : 55.542 85.463 47.069 -> 85.463 -> 2051.121 MByte/s p06 random-cyc-1dim : 39.881 45.733 35.412 -> 45.733 -> 1097.590 MByte/s p07 random-cyc-1dim : 44.061 45.451 35.149 -> 45.451 -> 1090.812 MByte/s p08 random-cyc-1dim : 41.507 47.110 36.829 -> 47.110 -> 1130.635 MByte/s p09 random-cyc-1dim : 45.395 48.393 38.475 -> 48.393 -> 1161.433 MByte/s p10 random-cyc-1dim : 39.785 47.700 34.997 -> 47.700 -> 1144.797 MByte/s p11 random-cyc-1dim : 37.739 46.771 35.835 -> 46.771 -> 1122.513 MByte/s p12 random-cyc-1dim : 38.003 48.623 35.611 -> 48.623 -> 1166.943 MByte/s p13 random-cyc-1dim : 40.401 43.620 35.889 -> 43.620 -> 1046.886 MByte/s p14 random-cyc-1dim : 39.026 45.234 35.207 -> 45.234 -> 1085.609 MByte/s p15 random-cyc-1dim : 41.997 48.855 36.717 -> 48.855 -> 1172.527 MByte/s -- additional patterns that are not used to compute b_eff: p16 worst-cyc-1dim : 36.540 44.517 32.767 -> 44.517 -> 1068.419 MByte/s p17 best bi-section : 51.752 80.139 49.837 -> 80.139 -> 1923.331 MByte/s p18 worst bi-section : 27.999 34.446 26.901 -> 34.446 -> 826.709 MByte/s p19 acyclic-1dim-all : 54.201 70.721 46.601 -> 70.721 -> 1697.306 MByte/s p20 acyclic-2dim-all : 42.719 48.592 35.736 -> 48.592 -> 1166.198 MByte/s p21 acyclic-3dim-all : 40.774 46.283 36.033 -> 46.283 -> 1110.788 MByte/s p22 cyclic-1dim-all : 56.167 84.932 47.170 -> 84.932 -> 2038.363 MByte/s p23 cyclic-2dim-all : 46.306 51.477 37.342 -> 51.477 -> 1235.444 MByte/s p24 cyclic-3dim-all : 44.815 51.817 38.297 -> 51.817 -> 1243.612 MByte/s log_avg of all rings : 59.010 81.930 49.013 || 81.930 -> 1966.329 MByte/s log_avg of all random : 40.712 46.720 35.998 || 46.720 -> 1121.288 MByte/s log_avg(ring,random) : 49.015 61.869 42.005 ||( 61.869 -> 1484.864)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-12*2fix : 80.112 79.705 80.269 -> 80.269 -> 1926.466 MByte/s p01 ring-6*4fix : 81.084 80.951 81.779 -> 81.779 -> 1962.695 MByte/s p02 ring-3*8fix : 73.291 71.826 72.057 -> 73.291 -> 1758.978 MByte/s p03 ring-1*24fix : 84.435 84.690 83.012 -> 84.690 -> 2032.561 MByte/s p04 ring-1*24fix : 84.710 84.830 84.090 -> 84.830 -> 2035.920 MByte/s p05 ring-1*24fix : 84.123 84.859 84.185 -> 84.859 -> 2036.623 MByte/s p06 random-cyc-1dim : 45.853 45.743 46.374 -> 46.374 -> 1112.976 MByte/s p07 random-cyc-1dim : 49.873 50.672 49.295 -> 50.672 -> 1216.127 MByte/s p08 random-cyc-1dim : 49.006 49.788 49.549 -> 49.788 -> 1194.901 MByte/s p09 random-cyc-1dim : 52.525 52.564 51.675 -> 52.564 -> 1261.533 MByte/s p10 random-cyc-1dim : 47.606 47.517 47.367 -> 47.606 -> 1142.533 MByte/s p11 random-cyc-1dim : 46.399 46.229 46.232 -> 46.399 -> 1113.568 MByte/s p12 random-cyc-1dim : 47.563 48.015 47.053 -> 48.015 -> 1152.350 MByte/s p13 random-cyc-1dim : 46.476 45.866 47.511 -> 47.511 -> 1140.254 MByte/s p14 random-cyc-1dim : 46.055 45.555 45.662 -> 46.055 -> 1105.312 MByte/s p15 random-cyc-1dim : 49.949 49.454 50.510 -> 50.510 -> 1212.240 MByte/s -- additional patterns that are not used to compute b_eff: p16 worst-cyc-1dim : 43.703 43.764 43.621 -> 43.764 -> 1050.324 MByte/s p17 best bi-section : 79.335 79.226 79.529 -> 79.529 -> 1908.700 MByte/s p18 worst bi-section : 34.508 34.329 34.386 -> 34.508 -> 828.202 MByte/s p19 acyclic-1dim-all : 69.190 70.019 69.832 -> 70.019 -> 1680.455 MByte/s p20 acyclic-2dim-all : 52.624 52.570 52.837 -> 52.837 -> 1268.078 MByte/s p21 acyclic-3dim-all : 51.485 50.474 51.011 -> 51.485 -> 1235.646 MByte/s p22 cyclic-1dim-all : 84.249 84.051 84.203 -> 84.249 -> 2021.987 MByte/s p23 cyclic-2dim-all : 55.440 54.801 54.880 -> 55.440 -> 1330.557 MByte/s p24 cyclic-3dim-all : 54.788 54.429 54.294 -> 54.788 -> 1314.923 MByte/s log_avg of all rings : 81.191 81.005 80.785 || 81.512 -> 1956.282 MByte/s log_avg of all random : 48.087 48.087 48.085 || 48.505 -> 1164.110 MByte/s log_avg(ring,random) : 62.484 62.412 62.326 ||( 62.878 -> 1509.082)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 65536 1048576 -> average -> accumulated p00 ring-12*2fix p00 method 0 : 0.055 0.887 10.513 51.347 178.993 208.577 -> 72.290 -> 1734.970 MByte/s p00 method 1 : 0.033 0.595 8.990 86.767 192.804 209.316 -> 80.193 -> 1924.632 MByte/s p00 method 2 : 0.025 0.405 5.337 26.647 169.351 163.670 -> 61.217 -> 1469.207 MByte/s p01 ring-6*4fix p01 method 0 : 0.055 0.858 9.949 49.795 165.041 154.090 -> 61.416 -> 1473.996 MByte/s p01 method 1 : 0.049 0.890 12.760 103.454 191.705 195.497 -> 82.282 -> 1974.771 MByte/s p01 method 2 : 0.023 0.347 4.723 29.600 119.485 112.662 -> 46.577 -> 1117.858 MByte/s p02 ring-3*8fix p02 method 0 : 0.053 0.835 9.616 49.708 135.316 140.604 -> 53.921 -> 1294.092 MByte/s p02 method 1 : 0.049 0.886 12.763 102.858 154.022 186.877 -> 73.696 -> 1768.702 MByte/s p02 method 2 : 0.022 0.343 4.528 30.110 119.041 122.047 -> 45.826 -> 1099.821 MByte/s p03 ring-1*24fix p03 method 0 : 0.049 0.783 8.441 42.430 144.919 144.712 -> 56.418 -> 1354.036 MByte/s p03 method 1 : 0.047 0.867 12.425 100.828 199.823 212.656 -> 85.382 -> 2049.157 MByte/s p03 method 2 : 0.022 0.344 4.493 30.178 126.006 124.147 -> 47.606 -> 1142.553 MByte/s p04 ring-1*24fix p04 method 0 : 0.049 0.782 8.495 43.087 144.860 131.342 -> 56.286 -> 1350.873 MByte/s p04 method 1 : 0.047 0.868 12.466 98.279 199.693 212.543 -> 85.239 -> 2045.745 MByte/s p04 method 2 : 0.022 0.345 4.489 30.067 126.852 120.482 -> 47.351 -> 1136.422 MByte/s p05 ring-1*24fix p05 method 0 : 0.048 0.784 8.399 45.671 143.705 133.268 -> 55.542 -> 1333.012 MByte/s p05 method 1 : 0.048 0.866 12.421 100.160 199.094 214.059 -> 85.463 -> 2051.121 MByte/s p05 method 2 : 0.022 0.343 4.464 29.953 128.988 112.582 -> 47.069 -> 1129.661 MByte/s p06 random-cyc-1dim p06 method 0 : 0.047 0.756 8.677 38.614 98.569 87.544 -> 39.881 -> 957.139 MByte/s p06 method 1 : 0.048 0.873 12.572 76.512 100.757 88.029 -> 45.733 -> 1097.590 MByte/s p06 method 2 : 0.021 0.342 4.433 29.510 86.261 84.888 -> 35.412 -> 849.886 MByte/s p07 random-cyc-1dim p07 method 0 : 0.046 0.754 8.026 39.679 106.948 111.182 -> 44.061 -> 1057.469 MByte/s p07 method 1 : 0.047 0.849 12.348 81.309 97.399 74.505 -> 45.451 -> 1090.812 MByte/s p07 method 2 : 0.021 0.342 4.405 29.521 88.867 81.201 -> 35.149 -> 843.579 MByte/s p08 random-cyc-1dim p08 method 0 : 0.047 0.737 7.940 37.852 102.559 92.080 -> 41.507 -> 996.177 MByte/s p08 method 1 : 0.048 0.868 12.481 85.355 104.873 70.429 -> 47.110 -> 1130.635 MByte/s p08 method 2 : 0.021 0.338 4.326 29.398 93.648 96.747 -> 36.829 -> 883.900 MByte/s p09 random-cyc-1dim p09 method 0 : 0.047 0.782 8.089 40.817 113.328 118.925 -> 45.395 -> 1089.477 MByte/s p09 method 1 : 0.049 0.860 12.475 87.871 101.747 82.518 -> 48.393 -> 1161.433 MByte/s p09 method 2 : 0.022 0.343 4.415 29.466 96.613 96.411 -> 38.475 -> 923.411 MByte/s p10 random-cyc-1dim p10 method 0 : 0.047 0.755 8.210 37.958 95.705 102.081 -> 39.785 -> 954.843 MByte/s p10 method 1 : 0.048 0.852 12.378 76.347 100.004 94.916 -> 47.700 -> 1144.797 MByte/s p10 method 2 : 0.021 0.338 4.395 29.542 88.893 81.408 -> 34.997 -> 839.934 MByte/s p11 random-cyc-1dim p11 method 0 : 0.047 0.777 8.286 36.874 92.707 77.187 -> 37.739 -> 905.745 MByte/s p11 method 1 : 0.049 0.873 12.579 76.820 99.770 89.171 -> 46.771 -> 1122.513 MByte/s p11 method 2 : 0.022 0.341 4.428 29.537 89.464 89.221 -> 35.835 -> 860.052 MByte/s p12 random-cyc-1dim p12 method 0 : 0.046 0.735 7.984 39.362 92.763 79.216 -> 38.003 -> 912.084 MByte/s p12 method 1 : 0.049 0.862 12.447 82.441 104.394 90.869 -> 48.623 -> 1166.943 MByte/s p12 method 2 : 0.021 0.338 4.387 29.339 85.296 90.931 -> 35.611 -> 854.656 MByte/s p13 random-cyc-1dim p13 method 0 : 0.046 0.738 7.863 38.597 96.234 103.115 -> 40.401 -> 969.629 MByte/s p13 method 1 : 0.049 0.867 12.500 77.645 92.177 64.714 -> 43.620 -> 1046.886 MByte/s p13 method 2 : 0.021 0.338 4.411 29.391 86.915 105.069 -> 35.889 -> 861.330 MByte/s p14 random-cyc-1dim p14 method 0 : 0.046 0.768 7.970 38.680 97.709 88.335 -> 39.026 -> 936.628 MByte/s p14 method 1 : 0.050 0.876 12.632 80.920 94.458 79.994 -> 45.234 -> 1085.609 MByte/s p14 method 2 : 0.021 0.340 4.365 29.443 86.440 82.479 -> 35.207 -> 844.963 MByte/s p15 random-cyc-1dim p15 method 0 : 0.046 0.728 7.835 38.334 106.023 97.941 -> 41.997 -> 1007.938 MByte/s p15 method 1 : 0.049 0.865 12.530 87.781 103.632 78.371 -> 48.855 -> 1172.527 MByte/s p15 method 2 : 0.021 0.340 4.349 29.505 93.349 89.393 -> 36.717 -> 881.199 MByte/s -- additional patterns that are not used to compute b_eff: p16 worst-cyc-1dim p16 method 0 : 0.045 0.717 7.864 38.786 87.796 83.173 -> 36.540 -> 876.959 MByte/s p16 method 1 : 0.049 0.862 12.427 78.348 88.495 89.555 -> 44.517 -> 1068.419 MByte/s p16 method 2 : 0.021 0.331 4.295 29.212 81.091 81.737 -> 32.767 -> 786.411 MByte/s p17 best bi-section p17 method 0 : 0.034 0.512 5.632 28.364 130.593 163.217 -> 51.752 -> 1242.056 MByte/s p17 method 1 : 0.034 0.587 8.709 85.901 193.478 209.411 -> 80.139 -> 1923.331 MByte/s p17 method 2 : 0.015 0.238 3.513 27.633 127.212 159.656 -> 49.837 -> 1196.082 MByte/s p18 worst bi-section p18 method 0 : 0.024 0.377 4.623 26.692 66.159 74.895 -> 27.999 -> 671.978 MByte/s p18 method 1 : 0.032 0.561 8.335 57.143 64.148 87.337 -> 34.446 -> 826.709 MByte/s p18 method 2 : 0.015 0.238 3.429 26.924 64.892 56.652 -> 26.901 -> 645.624 MByte/s p19 acyclic-1dim-all p19 method 0 : 0.047 0.763 8.551 41.687 142.258 128.385 -> 54.201 -> 1300.816 MByte/s p19 method 1 : 0.049 0.859 12.303 100.232 166.427 141.818 -> 70.721 -> 1697.306 MByte/s p19 method 2 : 0.021 0.330 4.318 29.455 127.509 120.152 -> 46.601 -> 1118.413 MByte/s p20 acyclic-2dim-all p20 method 0 : 0.034 0.544 6.006 27.108 108.263 122.700 -> 42.719 -> 1025.262 MByte/s p20 method 1 : 0.054 0.962 13.312 83.548 107.024 83.842 -> 48.592 -> 1166.198 MByte/s p20 method 2 : 0.017 0.274 3.561 24.336 94.844 88.017 -> 35.736 -> 857.660 MByte/s p21 acyclic-3dim-all p21 method 0 : 0.030 0.470 5.308 25.493 103.707 118.123 -> 40.774 -> 978.564 MByte/s p21 method 1 : 0.056 1.014 13.580 78.434 103.224 65.744 -> 46.283 -> 1110.788 MByte/s p21 method 2 : 0.015 0.240 3.099 21.867 96.031 90.464 -> 36.033 -> 864.785 MByte/s p22 cyclic-1dim-all p22 method 0 : 0.048 0.772 8.520 41.263 143.760 144.763 -> 56.167 -> 1348.012 MByte/s p22 method 1 : 0.046 0.824 11.983 98.598 198.847 213.988 -> 84.932 -> 2038.363 MByte/s p22 method 2 : 0.022 0.347 4.543 29.863 124.994 112.708 -> 47.170 -> 1132.075 MByte/s p23 cyclic-2dim-all p23 method 0 : 0.040 0.713 7.111 35.382 117.230 123.672 -> 46.306 -> 1111.340 MByte/s p23 method 1 : 0.062 1.109 15.371 85.843 106.660 91.278 -> 51.477 -> 1235.444 MByte/s p23 method 2 : 0.022 0.344 4.434 29.949 95.391 81.280 -> 37.342 -> 896.217 MByte/s p24 cyclic-3dim-all p24 method 0 : 0.043 0.657 6.973 32.332 111.538 119.816 -> 44.815 -> 1075.558 MByte/s p24 method 1 : 0.066 1.204 16.475 85.040 105.836 95.913 -> 51.817 -> 1243.612 MByte/s p24 method 2 : 0.022 0.340 4.385 29.762 97.068 91.178 -> 38.297 -> 919.122 MByte/s log_avg of all rings - ring, method 0 : 0.051 0.820 9.198 46.877 151.433 150.122 || 59.010 -> 1416.246 MByte/s - ring, method 1 : 0.045 0.821 11.884 98.555 188.756 204.892 || 81.930 -> 1966.329 MByte/s - ring, method 2 : 0.022 0.354 4.663 29.398 130.619 124.863 || 49.013 -> 1176.321 MByte/s log_avg of all random - random, method 0 : 0.046 0.753 8.085 38.663 100.053 94.916 || 40.712 -> 977.097 MByte/s - random, method 1 : 0.049 0.864 12.494 81.188 99.841 80.822 || 46.720 -> 1121.288 MByte/s - random, method 2 : 0.021 0.340 4.391 29.465 89.505 89.478 || 35.998 -> 863.958 MByte/s log_avg(ring,random) - average, method 0 : 0.049 0.786 8.624 42.572 123.091 119.370 || 49.015 -> 1176.354 MByte/s - average, method 1 : 0.047 0.842 12.185 89.451 137.279 128.685 || 61.869 -> 1484.864 MByte/s - average, method 2 : 0.022 0.347 4.525 29.431 108.125 105.700 || 42.005 -> 1008.113 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 1.199 0.050 0.051 0.049 0.049 0.047 0.022 2 2.398 0.100 0.103 0.097 0.098 0.094 0.044 4 5.415 0.226 0.227 0.225 0.195 0.218 0.088 8 10.739 0.447 0.450 0.445 0.390 0.432 0.176 16 20.899 0.871 0.877 0.864 0.786 0.842 0.347 32 40.721 1.697 1.688 1.706 1.217 1.666 0.606 64 74.926 3.122 3.104 3.140 2.192 3.080 1.158 128 159.723 6.655 6.591 6.719 4.741 6.551 2.394 256 296.278 12.345 12.198 12.494 8.624 12.185 4.525 512 545.012 22.709 22.429 22.992 15.447 22.560 8.242 1024 967.533 40.314 40.099 40.530 23.810 40.314 14.246 2048 1523.942 63.498 65.602 61.461 33.053 63.498 21.912 4096 2146.828 89.451 98.555 81.188 42.572 89.451 29.431 8192 2685.557 111.898 134.589 93.033 67.063 111.898 63.791 16384 3043.021 126.793 163.336 98.425 91.976 126.793 86.229 32768 3207.041 133.627 180.006 99.197 111.394 133.486 100.711 65536 3344.656 139.361 188.756 102.892 123.091 137.279 108.125 131072 3396.886 141.537 193.975 103.275 127.820 136.760 111.090 262144 3427.539 142.814 198.193 102.909 128.319 136.068 112.665 524288 3422.002 142.583 203.614 99.846 124.159 133.315 108.180 1048576 3407.119 141.963 204.892 98.362 119.370 128.685 105.700 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 65536 1048576 -> average -> accumulated p00 ring-12*2fix : 0.055 0.887 10.513 86.767 192.804 209.316 -> 80.438 -> 1930.518 MByte/s p01 ring-6*4fix : 0.055 0.890 12.760 103.454 191.705 195.497 -> 82.283 -> 1974.789 MByte/s p02 ring-3*8fix : 0.053 0.886 12.763 102.858 154.022 186.877 -> 73.696 -> 1768.714 MByte/s p03 ring-1*24fix : 0.049 0.867 12.425 100.828 199.823 212.656 -> 85.382 -> 2049.164 MByte/s p04 ring-1*24fix : 0.049 0.868 12.466 98.279 199.693 212.543 -> 85.240 -> 2045.748 MByte/s p05 ring-1*24fix : 0.048 0.866 12.421 100.160 199.094 214.059 -> 85.463 -> 2051.124 MByte/s p06 random-cyc-1dim : 0.048 0.873 12.572 76.512 100.757 88.029 -> 46.908 -> 1125.791 MByte/s p07 random-cyc-1dim : 0.047 0.849 12.348 81.309 106.948 111.182 -> 51.114 -> 1226.726 MByte/s p08 random-cyc-1dim : 0.048 0.868 12.481 85.355 104.873 96.747 -> 50.385 -> 1209.251 MByte/s p09 random-cyc-1dim : 0.049 0.860 12.475 87.871 113.328 118.925 -> 53.876 -> 1293.013 MByte/s p10 random-cyc-1dim : 0.048 0.852 12.378 76.347 100.004 102.081 -> 48.207 -> 1156.957 MByte/s p11 random-cyc-1dim : 0.049 0.873 12.579 76.820 99.770 89.221 -> 46.774 -> 1122.570 MByte/s p12 random-cyc-1dim : 0.049 0.862 12.447 82.441 104.394 90.931 -> 48.626 -> 1167.014 MByte/s p13 random-cyc-1dim : 0.049 0.867 12.500 77.645 96.234 105.069 -> 47.933 -> 1150.385 MByte/s p14 random-cyc-1dim : 0.050 0.876 12.632 80.920 97.709 88.335 -> 46.585 -> 1118.033 MByte/s p15 random-cyc-1dim : 0.049 0.865 12.530 87.781 106.023 97.941 -> 50.857 -> 1220.569 MByte/s -- additional patterns that are not used to compute b_eff: p16 worst-cyc-1dim : 0.049 0.862 12.427 78.348 88.495 89.555 -> 44.517 -> 1068.419 MByte/s p17 best bi-section : 0.034 0.587 8.709 85.901 193.478 209.411 -> 80.139 -> 1923.332 MByte/s p18 worst bi-section : 0.032 0.561 8.335 57.143 66.159 87.337 -> 34.640 -> 831.367 MByte/s p19 acyclic-1dim-all : 0.049 0.859 12.303 100.232 166.427 141.818 -> 70.721 -> 1697.306 MByte/s p20 acyclic-2dim-all : 0.054 0.962 13.312 83.548 108.263 122.700 -> 53.597 -> 1286.322 MByte/s p21 acyclic-3dim-all : 0.056 1.014 13.580 78.434 103.707 118.123 -> 51.786 -> 1242.866 MByte/s p22 cyclic-1dim-all : 0.048 0.824 11.983 98.598 198.847 213.988 -> 84.932 -> 2038.372 MByte/s p23 cyclic-2dim-all : 0.062 1.109 15.371 85.843 117.230 123.672 -> 55.940 -> 1342.553 MByte/s p24 cyclic-3dim-all : 0.066 1.204 16.475 85.040 111.538 119.816 -> 55.272 -> 1326.523 MByte/s log_avg of all rings : 0.051 0.877 12.198 98.555 188.756 204.892 || 81.972 -> 1967.337 MByte/s log_avg of all random : 0.049 0.864 12.494 81.188 102.892 98.362 || 49.076 -> 1177.821 MByte/s log_avg(ring,random) : 0.050 0.871 12.345 89.451 139.361 141.963 || 63.426 -> 1522.225 MByte/s SECTION-BY-PATTERN-MSGLNG-END SECTION-BEFF-BEGIN The effective bandwidth is b_eff = 1522.225 MByte/s on 24 processes ( = 63.426 MByte/s * 24 processes) system parameters : 24 nodes, 128 MB/node system name: sn6715 hostname : hwwt3e OS release : 2.0.4.71 OS version : unicosmk machine : CRAY T3E SECTION-BEFF-END b_eff = 1522.225 MB/s = 63.426 * 24 PEs with 128 MB/PE on sn6715 hwwt3e 2.0.4.71 unicosmk CRAY T3E