b_eff = 1063.217 MB/s = 66.451 * 16 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 = 16 1-dim-paterns: size = 16 2-dim-paterns: size = 4 * 4 3-dim-paterns: size = 4 * 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 (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 = 124.763 sec sum of max elapsed time per entries above = 126.778 sec difference = -2.016 sec = 1.6% The difference is less than 5 % SECTION-LOOP-END SECTION-PATTERN-BEGIN Pattern parameters: p00 ring-8*2fix => 1 sendrecv_calls with 16 messages, i.e. msgs/used node, all nodes are used p01 ring-4*4fix => 2 sendrecv_calls with 32 messages, i.e. msgs/used node, all nodes are used p02 ring-2*8fix => 2 sendrecv_calls with 32 messages, i.e. msgs/used node, all nodes are used p03 ring-1*16fix => 2 sendrecv_calls with 32 messages, i.e. msgs/used node, all nodes are used p04 ring-1*16fix => 2 sendrecv_calls with 32 messages, i.e. msgs/used node, all nodes are used p05 ring-1*16fix => 2 sendrecv_calls with 32 messages, i.e. msgs/used node, all nodes are used p06 random-cyc-1dim => 2 sendrecv_calls with 32 messages, i.e. msgs/used node, all nodes are used p07 random-cyc-1dim => 2 sendrecv_calls with 32 messages, i.e. msgs/used node, all nodes are used p08 random-cyc-1dim => 2 sendrecv_calls with 32 messages, i.e. msgs/used node, all nodes are used p09 random-cyc-1dim => 2 sendrecv_calls with 32 messages, i.e. msgs/used node, all nodes are used p10 random-cyc-1dim => 2 sendrecv_calls with 32 messages, i.e. msgs/used node, all nodes are used p11 random-cyc-1dim => 2 sendrecv_calls with 32 messages, i.e. msgs/used node, all nodes are used p12 random-cyc-1dim => 2 sendrecv_calls with 32 messages, i.e. msgs/used node, all nodes are used p13 random-cyc-1dim => 2 sendrecv_calls with 32 messages, i.e. msgs/used node, all nodes are used p14 random-cyc-1dim => 2 sendrecv_calls with 32 messages, i.e. msgs/used node, all nodes are used p15 random-cyc-1dim => 2 sendrecv_calls with 32 messages, i.e. msgs/used node, all nodes are used p16 worst-cyc-1dim => 2 sendrecv_calls with 32 messages, i.e. msgs/used node, all nodes are used p17 best bi-section => 2 sendrecv_calls with 16 messages, i.e. msgs/used node, all nodes are used p18 worst bi-section => 2 sendrecv_calls with 16 messages, i.e. msgs/used node, all nodes are used p19 acyclic-1dim-all => 2 sendrecv_calls with 30 messages, i.e. msgs/used node, all nodes are used p20 acyclic-2dim-all => 4 sendrecv_calls with 48 messages, i.e. msgs/used node, all nodes are used p21 acyclic-3dim-all => 6 sendrecv_calls with 56 messages, i.e. msgs/used node, all nodes are used p22 cyclic-1dim-all => 2 sendrecv_calls with 32 messages, i.e. msgs/used node, all nodes are used p23 cyclic-2dim-all => 4 sendrecv_calls with 64 messages, i.e. msgs/used node, all nodes are used p24 cyclic-3dim-all => 4 sendrecv_calls with 64 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-8*2fix : 72.339 83.797 65.461 -> 83.797 -> 1340.745 MByte/s p01 ring-4*4fix : 63.341 84.825 47.230 -> 84.825 -> 1357.198 MByte/s p02 ring-2*8fix : 55.125 75.879 46.054 -> 75.879 -> 1214.061 MByte/s p03 ring-1*16fix : 53.447 73.321 45.252 -> 73.321 -> 1173.144 MByte/s p04 ring-1*16fix : 52.940 73.296 44.202 -> 73.296 -> 1172.733 MByte/s p05 ring-1*16fix : 53.022 74.322 44.985 -> 74.322 -> 1189.149 MByte/s p06 random-cyc-1dim : 49.091 58.401 41.213 -> 58.401 -> 934.414 MByte/s p07 random-cyc-1dim : 43.074 46.268 36.630 -> 46.268 -> 740.283 MByte/s p08 random-cyc-1dim : 46.299 57.538 39.255 -> 57.538 -> 920.605 MByte/s p09 random-cyc-1dim : 46.202 53.629 41.145 -> 53.629 -> 858.059 MByte/s p10 random-cyc-1dim : 45.979 51.750 39.067 -> 51.750 -> 828.000 MByte/s p11 random-cyc-1dim : 44.588 51.253 37.963 -> 51.253 -> 820.050 MByte/s p12 random-cyc-1dim : 41.104 46.592 36.296 -> 46.592 -> 745.474 MByte/s p13 random-cyc-1dim : 51.400 62.456 41.768 -> 62.456 -> 999.294 MByte/s p14 random-cyc-1dim : 52.218 56.482 43.073 -> 56.482 -> 903.713 MByte/s p15 random-cyc-1dim : 47.335 57.142 39.473 -> 57.142 -> 914.277 MByte/s -- additional patterns that are not used to compute b_eff: p16 worst-cyc-1dim : 31.739 37.285 27.997 -> 37.285 -> 596.565 MByte/s p17 best bi-section : 51.725 83.675 49.915 -> 83.675 -> 1338.795 MByte/s p18 worst bi-section : 26.686 35.508 26.627 -> 35.508 -> 568.125 MByte/s p19 acyclic-1dim-all : 54.855 71.524 45.359 -> 71.524 -> 1144.379 MByte/s p20 acyclic-2dim-all : 43.370 48.687 35.112 -> 48.687 -> 778.990 MByte/s p21 acyclic-3dim-all : 42.105 53.732 39.060 -> 53.732 -> 859.717 MByte/s p22 cyclic-1dim-all : 53.291 72.914 45.994 -> 72.914 -> 1166.621 MByte/s p23 cyclic-2dim-all : 45.911 53.761 37.555 -> 53.761 -> 860.181 MByte/s p24 cyclic-3dim-all : 46.598 53.526 39.898 -> 53.526 -> 856.419 MByte/s log_avg of all rings : 57.955 77.425 48.373 || 77.425 -> 1238.804 MByte/s log_avg of all random : 46.613 53.919 39.532 || 53.919 -> 862.707 MByte/s log_avg(ring,random) : 51.976 64.612 43.730 ||( 64.612 -> 1033.791)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-8*2fix : 82.848 83.240 83.049 -> 83.240 -> 1331.845 MByte/s p01 ring-4*4fix : 82.573 84.337 82.639 -> 84.337 -> 1349.392 MByte/s p02 ring-2*8fix : 74.907 74.324 75.126 -> 75.126 -> 1202.019 MByte/s p03 ring-1*16fix : 71.533 72.673 71.602 -> 72.673 -> 1162.774 MByte/s p04 ring-1*16fix : 72.540 72.153 71.815 -> 72.540 -> 1160.645 MByte/s p05 ring-1*16fix : 72.553 72.057 72.010 -> 72.553 -> 1160.847 MByte/s p06 random-cyc-1dim : 59.684 59.945 59.863 -> 59.945 -> 959.117 MByte/s p07 random-cyc-1dim : 50.767 50.943 50.198 -> 50.943 -> 815.095 MByte/s p08 random-cyc-1dim : 57.043 57.316 57.262 -> 57.316 -> 917.055 MByte/s p09 random-cyc-1dim : 55.821 54.938 54.408 -> 55.821 -> 893.136 MByte/s p10 random-cyc-1dim : 54.404 54.583 54.522 -> 54.583 -> 873.332 MByte/s p11 random-cyc-1dim : 53.667 53.303 53.891 -> 53.891 -> 862.264 MByte/s p12 random-cyc-1dim : 49.922 50.357 50.345 -> 50.357 -> 805.709 MByte/s p13 random-cyc-1dim : 61.396 61.417 61.770 -> 61.770 -> 988.316 MByte/s p14 random-cyc-1dim : 59.712 60.865 58.693 -> 60.865 -> 973.836 MByte/s p15 random-cyc-1dim : 56.956 57.275 58.158 -> 58.158 -> 930.524 MByte/s -- additional patterns that are not used to compute b_eff: p16 worst-cyc-1dim : 37.686 37.362 37.741 -> 37.741 -> 603.850 MByte/s p17 best bi-section : 83.507 83.499 82.890 -> 83.507 -> 1336.114 MByte/s p18 worst bi-section : 35.425 35.680 35.715 -> 35.715 -> 571.447 MByte/s p19 acyclic-1dim-all : 71.532 70.660 71.618 -> 71.618 -> 1145.889 MByte/s p20 acyclic-2dim-all : 53.768 53.016 53.438 -> 53.768 -> 860.294 MByte/s p21 acyclic-3dim-all : 54.284 55.611 54.657 -> 55.611 -> 889.769 MByte/s p22 cyclic-1dim-all : 72.162 71.491 70.654 -> 72.162 -> 1154.594 MByte/s p23 cyclic-2dim-all : 54.969 55.051 54.272 -> 55.051 -> 880.815 MByte/s p24 cyclic-3dim-all : 55.380 55.984 55.835 -> 55.984 -> 895.743 MByte/s log_avg of all rings : 76.015 76.289 75.882 || 76.582 -> 1225.308 MByte/s log_avg of all random : 55.819 55.969 55.788 || 56.238 -> 899.808 MByte/s log_avg(ring,random) : 65.139 65.344 65.064 ||( 65.626 -> 1050.020)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-8*2fix p00 method 0 : 0.056 0.886 10.499 52.315 179.942 207.980 -> 72.339 -> 1157.417 MByte/s p00 method 1 : 0.042 0.780 11.179 99.759 196.101 209.516 -> 83.797 -> 1340.745 MByte/s p00 method 2 : 0.025 0.401 5.351 26.392 169.165 206.933 -> 65.461 -> 1047.383 MByte/s p01 ring-4*4fix p01 method 0 : 0.055 0.859 9.973 50.061 171.844 153.063 -> 63.341 -> 1013.459 MByte/s p01 method 1 : 0.059 1.082 15.225 111.882 193.172 197.850 -> 84.825 -> 1357.198 MByte/s p01 method 2 : 0.022 0.368 4.711 29.840 123.521 116.513 -> 47.230 -> 755.673 MByte/s p02 ring-2*8fix p02 method 0 : 0.052 0.842 9.554 49.969 137.167 141.168 -> 55.125 -> 882.007 MByte/s p02 method 1 : 0.058 1.061 14.927 110.671 156.609 189.859 -> 75.879 -> 1214.061 MByte/s p02 method 2 : 0.022 0.348 4.538 30.271 119.259 122.736 -> 46.054 -> 736.865 MByte/s p03 ring-1*16fix p03 method 0 : 0.049 0.787 8.645 44.560 133.635 146.606 -> 53.447 -> 855.152 MByte/s p03 method 1 : 0.056 1.051 14.588 104.531 161.501 177.623 -> 73.321 -> 1173.144 MByte/s p03 method 2 : 0.022 0.344 4.479 30.263 115.122 111.121 -> 45.252 -> 724.032 MByte/s p04 ring-1*16fix p04 method 0 : 0.049 0.817 8.699 44.349 136.579 131.594 -> 52.940 -> 847.039 MByte/s p04 method 1 : 0.056 1.046 14.598 104.367 154.707 173.351 -> 73.296 -> 1172.733 MByte/s p04 method 2 : 0.022 0.344 4.557 30.170 117.740 112.220 -> 44.202 -> 707.229 MByte/s p05 ring-1*16fix p05 method 0 : 0.049 0.805 8.920 44.549 133.342 132.433 -> 53.022 -> 848.345 MByte/s p05 method 1 : 0.056 1.045 14.667 104.841 157.664 175.012 -> 74.322 -> 1189.149 MByte/s p05 method 2 : 0.022 0.345 4.484 30.286 116.442 113.950 -> 44.985 -> 719.766 MByte/s p06 random-cyc-1dim p06 method 0 : 0.051 0.838 9.239 44.608 119.921 124.290 -> 49.091 -> 785.458 MByte/s p06 method 1 : 0.057 1.057 14.742 102.621 127.399 98.564 -> 58.401 -> 934.414 MByte/s p06 method 2 : 0.022 0.345 4.405 29.699 104.751 98.610 -> 41.213 -> 659.412 MByte/s p07 random-cyc-1dim p07 method 0 : 0.048 0.779 8.362 40.407 104.953 98.744 -> 43.074 -> 689.186 MByte/s p07 method 1 : 0.058 1.069 14.921 84.518 97.957 70.705 -> 46.268 -> 740.283 MByte/s p07 method 2 : 0.021 0.345 4.466 29.483 93.608 99.164 -> 36.630 -> 586.080 MByte/s p08 random-cyc-1dim p08 method 0 : 0.049 0.796 8.861 44.777 112.654 111.079 -> 46.299 -> 740.777 MByte/s p08 method 1 : 0.057 1.057 14.706 95.649 119.734 117.238 -> 57.538 -> 920.605 MByte/s p08 method 2 : 0.022 0.343 4.439 29.862 101.770 93.161 -> 39.255 -> 628.078 MByte/s p09 random-cyc-1dim p09 method 0 : 0.047 0.783 8.741 41.066 117.177 110.908 -> 46.202 -> 739.237 MByte/s p09 method 1 : 0.058 1.058 14.970 95.400 113.176 96.579 -> 53.629 -> 858.059 MByte/s p09 method 2 : 0.021 0.343 4.468 29.852 100.635 114.170 -> 41.145 -> 658.317 MByte/s p10 random-cyc-1dim p10 method 0 : 0.047 0.777 8.796 39.015 113.440 111.814 -> 45.979 -> 735.659 MByte/s p10 method 1 : 0.059 1.064 15.091 94.111 112.140 93.829 -> 51.750 -> 828.000 MByte/s p10 method 2 : 0.021 0.343 4.452 29.791 100.113 96.844 -> 39.067 -> 625.075 MByte/s p11 random-cyc-1dim p11 method 0 : 0.049 0.791 8.636 44.276 106.854 105.076 -> 44.588 -> 713.413 MByte/s p11 method 1 : 0.056 1.009 14.329 93.750 107.230 85.254 -> 51.253 -> 820.050 MByte/s p11 method 2 : 0.022 0.345 4.482 29.870 97.251 104.334 -> 37.963 -> 607.414 MByte/s p12 random-cyc-1dim p12 method 0 : 0.047 0.759 8.396 40.284 106.727 92.123 -> 41.104 -> 657.659 MByte/s p12 method 1 : 0.058 1.053 14.890 94.934 100.037 62.847 -> 46.592 -> 745.474 MByte/s p12 method 2 : 0.021 0.344 4.427 29.849 88.135 88.133 -> 36.296 -> 580.737 MByte/s p13 random-cyc-1dim p13 method 0 : 0.050 0.801 8.785 46.362 126.538 127.640 -> 51.400 -> 822.399 MByte/s p13 method 1 : 0.058 1.048 14.662 98.307 126.257 134.467 -> 62.456 -> 999.294 MByte/s p13 method 2 : 0.022 0.346 4.488 30.003 113.930 103.495 -> 41.768 -> 668.281 MByte/s p14 random-cyc-1dim p14 method 0 : 0.050 0.796 8.836 41.616 129.637 137.439 -> 52.218 -> 835.482 MByte/s p14 method 1 : 0.059 1.067 15.081 97.343 112.579 127.414 -> 56.482 -> 903.713 MByte/s p14 method 2 : 0.022 0.345 4.515 29.985 110.275 115.686 -> 43.073 -> 689.172 MByte/s p15 random-cyc-1dim p15 method 0 : 0.050 0.795 8.585 40.247 113.716 129.820 -> 47.335 -> 757.367 MByte/s p15 method 1 : 0.059 1.063 14.900 100.453 118.270 108.116 -> 57.142 -> 914.277 MByte/s p15 method 2 : 0.022 0.342 4.472 29.950 100.484 105.232 -> 39.473 -> 631.570 MByte/s -- additional patterns that are not used to compute b_eff: p16 worst-cyc-1dim p16 method 0 : 0.047 0.785 8.666 37.129 74.854 63.384 -> 31.739 -> 507.820 MByte/s p16 method 1 : 0.059 1.051 14.635 69.139 74.476 53.733 -> 37.285 -> 596.565 MByte/s p16 method 2 : 0.021 0.336 4.319 29.260 65.460 61.789 -> 27.997 -> 447.960 MByte/s p17 best bi-section p17 method 0 : 0.034 0.516 5.704 28.239 131.674 162.930 -> 51.725 -> 827.599 MByte/s p17 method 1 : 0.043 0.770 10.926 98.808 196.684 209.891 -> 83.675 -> 1338.795 MByte/s p17 method 2 : 0.015 0.238 3.493 27.726 127.778 161.447 -> 49.915 -> 798.648 MByte/s p18 worst bi-section p18 method 0 : 0.025 0.383 4.701 26.745 65.552 55.801 -> 26.686 -> 426.979 MByte/s p18 method 1 : 0.041 0.727 10.528 59.108 63.873 87.498 -> 35.508 -> 568.125 MByte/s p18 method 2 : 0.015 0.237 3.501 26.913 68.084 63.820 -> 26.627 -> 426.028 MByte/s p19 acyclic-1dim-all p19 method 0 : 0.047 0.758 8.825 43.984 136.953 151.312 -> 54.855 -> 877.681 MByte/s p19 method 1 : 0.057 1.022 14.378 105.231 162.437 137.987 -> 71.524 -> 1144.379 MByte/s p19 method 2 : 0.020 0.322 4.206 29.107 120.502 115.473 -> 45.359 -> 725.751 MByte/s p20 acyclic-2dim-all p20 method 0 : 0.034 0.526 5.940 25.827 109.793 124.498 -> 43.370 -> 693.917 MByte/s p20 method 1 : 0.057 1.043 13.869 85.739 104.837 77.350 -> 48.687 -> 778.990 MByte/s p20 method 2 : 0.016 0.260 3.414 23.314 95.083 76.714 -> 35.112 -> 561.786 MByte/s p21 acyclic-3dim-all p21 method 0 : 0.030 0.469 5.245 24.841 106.467 127.122 -> 42.105 -> 673.682 MByte/s p21 method 1 : 0.065 1.192 15.506 87.921 114.579 104.011 -> 53.732 -> 859.717 MByte/s p21 method 2 : 0.015 0.245 3.260 24.375 102.124 103.087 -> 39.060 -> 624.968 MByte/s p22 cyclic-1dim-all p22 method 0 : 0.049 0.800 8.993 47.125 134.656 131.962 -> 53.291 -> 852.661 MByte/s p22 method 1 : 0.055 0.994 13.970 101.497 153.781 173.896 -> 72.914 -> 1166.621 MByte/s p22 method 2 : 0.022 0.347 4.539 30.279 118.523 127.542 -> 45.994 -> 735.896 MByte/s p23 cyclic-2dim-all p23 method 0 : 0.052 0.820 9.714 34.984 112.385 118.505 -> 45.911 -> 734.570 MByte/s p23 method 1 : 0.069 1.272 17.138 89.109 109.467 101.317 -> 53.761 -> 860.181 MByte/s p23 method 2 : 0.022 0.341 4.461 30.035 94.981 86.367 -> 37.555 -> 600.887 MByte/s p24 cyclic-3dim-all p24 method 0 : 0.051 0.849 9.576 39.337 112.002 124.479 -> 46.598 -> 745.568 MByte/s p24 method 1 : 0.065 1.148 15.864 86.024 114.298 99.630 -> 53.526 -> 856.419 MByte/s p24 method 2 : 0.021 0.340 4.401 29.650 105.837 97.177 -> 39.898 -> 638.374 MByte/s log_avg of all rings - ring, method 0 : 0.051 0.832 9.357 47.524 147.558 150.200 || 57.955 -> 927.288 MByte/s - ring, method 1 : 0.054 1.005 14.123 105.929 169.087 186.744 || 77.425 -> 1238.804 MByte/s - ring, method 2 : 0.022 0.358 4.677 29.501 125.639 127.046 || 48.373 -> 773.967 MByte/s log_avg of all random - random, method 0 : 0.049 0.791 8.720 42.200 114.897 114.060 || 46.613 -> 745.809 MByte/s - random, method 1 : 0.058 1.054 14.827 95.594 113.084 96.966 || 53.919 -> 862.707 MByte/s - random, method 2 : 0.022 0.344 4.461 29.834 100.846 101.556 || 39.532 -> 632.509 MByte/s log_avg(ring,random) - average, method 0 : 0.050 0.811 9.033 44.783 130.208 130.889 || 51.976 -> 831.612 MByte/s - average, method 1 : 0.056 1.029 14.471 100.629 138.279 134.565 || 64.612 -> 1033.791 MByte/s - average, method 2 : 0.022 0.351 4.568 29.667 112.562 113.588 || 43.730 -> 699.672 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 0.917 0.057 0.057 0.058 0.050 0.056 0.022 2 1.846 0.115 0.115 0.116 0.101 0.112 0.044 4 4.348 0.272 0.267 0.276 0.201 0.269 0.088 8 8.689 0.543 0.534 0.553 0.404 0.537 0.176 16 16.643 1.040 1.026 1.054 0.811 1.029 0.351 32 31.797 1.987 1.937 2.039 1.293 1.987 0.612 64 58.743 3.671 3.602 3.742 2.302 3.671 1.163 128 126.925 7.933 7.750 8.120 5.043 7.933 2.416 256 231.535 14.471 14.123 14.827 9.033 14.471 4.568 512 421.422 26.339 25.850 26.837 15.885 26.339 8.332 1024 731.703 45.731 45.313 46.154 24.415 45.731 14.352 2048 1140.226 71.264 72.086 70.452 34.323 71.264 22.058 4096 1610.062 100.629 105.929 95.594 44.783 100.629 29.667 8192 1925.120 120.320 134.991 107.244 68.327 120.320 65.053 16384 2109.642 131.853 156.491 111.093 94.604 131.853 88.152 32768 2176.617 136.039 162.093 114.172 116.809 135.883 103.627 65536 2248.512 140.532 169.087 116.799 130.208 138.279 112.562 131072 2283.278 142.705 170.442 119.481 137.158 137.653 115.889 262144 2366.100 147.881 180.095 121.430 137.791 138.483 118.045 524288 2389.544 149.346 184.987 120.573 135.420 135.636 115.283 1048576 2351.447 146.965 186.744 115.660 130.889 134.565 113.588 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-8*2fix : 0.056 0.886 11.179 99.759 196.101 209.516 -> 83.808 -> 1340.924 MByte/s p01 ring-4*4fix : 0.059 1.082 15.225 111.882 193.172 197.850 -> 84.825 -> 1357.198 MByte/s p02 ring-2*8fix : 0.058 1.061 14.927 110.671 156.609 189.859 -> 75.879 -> 1214.061 MByte/s p03 ring-1*16fix : 0.056 1.051 14.588 104.531 161.501 177.623 -> 73.321 -> 1173.144 MByte/s p04 ring-1*16fix : 0.056 1.046 14.598 104.367 154.707 173.351 -> 73.296 -> 1172.733 MByte/s p05 ring-1*16fix : 0.056 1.045 14.667 104.841 157.664 175.012 -> 74.322 -> 1189.149 MByte/s p06 random-cyc-1dim : 0.057 1.057 14.742 102.621 127.399 124.290 -> 61.167 -> 978.679 MByte/s p07 random-cyc-1dim : 0.058 1.069 14.921 84.518 104.953 99.164 -> 51.493 -> 823.886 MByte/s p08 random-cyc-1dim : 0.057 1.057 14.706 95.649 119.734 117.238 -> 57.701 -> 923.214 MByte/s p09 random-cyc-1dim : 0.058 1.058 14.970 95.400 117.177 114.170 -> 56.397 -> 902.350 MByte/s p10 random-cyc-1dim : 0.059 1.064 15.091 94.111 113.440 111.814 -> 55.911 -> 894.573 MByte/s p11 random-cyc-1dim : 0.056 1.009 14.329 93.750 107.230 105.076 -> 54.325 -> 869.201 MByte/s p12 random-cyc-1dim : 0.058 1.053 14.890 94.934 106.727 92.123 -> 50.852 -> 813.631 MByte/s p13 random-cyc-1dim : 0.058 1.048 14.662 98.307 126.538 134.467 -> 63.142 -> 1010.274 MByte/s p14 random-cyc-1dim : 0.059 1.067 15.081 97.343 129.637 137.439 -> 61.892 -> 990.274 MByte/s p15 random-cyc-1dim : 0.059 1.063 14.900 100.453 118.270 129.820 -> 58.837 -> 941.393 MByte/s -- additional patterns that are not used to compute b_eff: p16 worst-cyc-1dim : 0.059 1.051 14.635 69.139 74.854 63.384 -> 38.072 -> 609.160 MByte/s p17 best bi-section : 0.043 0.770 10.926 98.808 196.684 209.891 -> 83.675 -> 1338.795 MByte/s p18 worst bi-section : 0.041 0.727 10.528 59.108 68.084 87.498 -> 35.822 -> 573.147 MByte/s p19 acyclic-1dim-all : 0.057 1.022 14.378 105.231 162.437 151.312 -> 72.158 -> 1154.531 MByte/s p20 acyclic-2dim-all : 0.057 1.043 13.869 85.739 109.793 124.498 -> 54.469 -> 871.512 MByte/s p21 acyclic-3dim-all : 0.065 1.192 15.506 87.921 114.579 127.122 -> 56.184 -> 898.947 MByte/s p22 cyclic-1dim-all : 0.055 0.994 13.970 101.497 153.781 173.896 -> 72.914 -> 1166.621 MByte/s p23 cyclic-2dim-all : 0.069 1.272 17.138 89.109 112.385 118.505 -> 55.987 -> 895.798 MByte/s p24 cyclic-3dim-all : 0.065 1.148 15.864 86.024 114.298 124.479 -> 56.443 -> 903.096 MByte/s log_avg of all rings : 0.057 1.026 14.123 105.929 169.087 186.744 || 77.427 -> 1238.832 MByte/s log_avg of all random : 0.058 1.054 14.827 95.594 116.799 115.660 || 57.031 -> 912.497 MByte/s log_avg(ring,random) : 0.057 1.040 14.471 100.629 140.532 146.965 || 66.451 -> 1063.217 MByte/s SECTION-BY-PATTERN-MSGLNG-END SECTION-BEFF-BEGIN The effective bandwidth is b_eff = 1063.217 MByte/s on 16 processes ( = 66.451 MByte/s * 16 processes) system parameters : 16 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 = 1063.217 MB/s = 66.451 * 16 PEs with 128 MB/PE on sn6715 hwwt3e 2.0.4.71 unicosmk CRAY T3E