b_eff = 3158.554 MB/s = 49.352 * 64 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 = 64 1-dim-paterns: size = 64 2-dim-paterns: size = 8 * 8 3-dim-paterns: size = 4 * 4 * 4 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 = 162.837 sec sum of max elapsed time per entries above = 166.771 sec difference = -3.934 sec = 2.4% The difference is less than 5 % SECTION-LOOP-END SECTION-PATTERN-BEGIN Pattern parameters: p00 ring-32*2fix => 1 sendrecv_calls with 64 messages, i.e. msgs/used node, all nodes are used p01 ring-16*4fix => 2 sendrecv_calls with 128 messages, i.e. msgs/used node, all nodes are used p02 ring-8*8fix => 2 sendrecv_calls with 128 messages, i.e. msgs/used node, all nodes are used p03 ring-4*16fix => 2 sendrecv_calls with 128 messages, i.e. msgs/used node, all nodes are used p04 ring-2*32fix => 2 sendrecv_calls with 128 messages, i.e. msgs/used node, all nodes are used p05 ring-1*64fix => 2 sendrecv_calls with 128 messages, i.e. msgs/used node, all nodes are used p06 random-cyc-1dim => 2 sendrecv_calls with 128 messages, i.e. msgs/used node, all nodes are used p07 random-cyc-1dim => 2 sendrecv_calls with 128 messages, i.e. msgs/used node, all nodes are used p08 random-cyc-1dim => 2 sendrecv_calls with 128 messages, i.e. msgs/used node, all nodes are used p09 random-cyc-1dim => 2 sendrecv_calls with 128 messages, i.e. msgs/used node, all nodes are used p10 random-cyc-1dim => 2 sendrecv_calls with 128 messages, i.e. msgs/used node, all nodes are used p11 random-cyc-1dim => 2 sendrecv_calls with 128 messages, i.e. msgs/used node, all nodes are used p12 random-cyc-1dim => 2 sendrecv_calls with 128 messages, i.e. msgs/used node, all nodes are used p13 random-cyc-1dim => 2 sendrecv_calls with 128 messages, i.e. msgs/used node, all nodes are used p14 random-cyc-1dim => 2 sendrecv_calls with 128 messages, i.e. msgs/used node, all nodes are used p15 random-cyc-1dim => 2 sendrecv_calls with 128 messages, i.e. msgs/used node, all nodes are used p16 worst-cyc-1dim => 2 sendrecv_calls with 128 messages, i.e. msgs/used node, all nodes are used p17 best bi-section => 2 sendrecv_calls with 64 messages, i.e. msgs/used node, all nodes are used p18 worst bi-section => 2 sendrecv_calls with 64 messages, i.e. msgs/used node, all nodes are used p19 acyclic-1dim-all => 2 sendrecv_calls with 126 messages, i.e. msgs/used node, all nodes are used p20 acyclic-2dim-all => 4 sendrecv_calls with 224 messages, i.e. msgs/used node, all nodes are used p21 acyclic-3dim-all => 6 sendrecv_calls with 288 messages, i.e. msgs/used node, all nodes are used p22 cyclic-1dim-all => 2 sendrecv_calls with 128 messages, i.e. msgs/used node, all nodes are used p23 cyclic-2dim-all => 4 sendrecv_calls with 256 messages, i.e. msgs/used node, all nodes are used p24 cyclic-3dim-all => 6 sendrecv_calls with 384 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-32*2fix : 72.164 70.847 60.632 -> 72.164 -> 4618.495 MByte/s p01 ring-16*4fix : 59.560 75.417 45.450 -> 75.417 -> 4826.718 MByte/s p02 ring-8*8fix : 53.474 67.317 44.340 -> 67.317 -> 4308.262 MByte/s p03 ring-4*16fix : 51.115 64.777 42.772 -> 64.777 -> 4145.708 MByte/s p04 ring-2*32fix : 54.760 78.580 46.037 -> 78.580 -> 5029.128 MByte/s p05 ring-1*64fix : 53.563 70.773 45.224 -> 70.773 -> 4529.489 MByte/s p06 random-cyc-1dim : 28.586 28.537 25.506 -> 28.586 -> 1829.512 MByte/s p07 random-cyc-1dim : 26.958 28.440 23.193 -> 28.440 -> 1820.155 MByte/s p08 random-cyc-1dim : 32.480 34.341 28.189 -> 34.341 -> 2197.816 MByte/s p09 random-cyc-1dim : 23.858 25.474 22.284 -> 25.474 -> 1630.312 MByte/s p10 random-cyc-1dim : 34.097 36.874 30.249 -> 36.874 -> 2359.913 MByte/s p11 random-cyc-1dim : 29.253 29.500 25.899 -> 29.500 -> 1887.985 MByte/s p12 random-cyc-1dim : 37.790 40.685 33.129 -> 40.685 -> 2603.867 MByte/s p13 random-cyc-1dim : 27.767 29.703 25.400 -> 29.703 -> 1901.005 MByte/s p14 random-cyc-1dim : 33.984 34.184 30.548 -> 34.184 -> 2187.785 MByte/s p15 random-cyc-1dim : 34.817 37.141 31.045 -> 37.141 -> 2377.018 MByte/s -- additional patterns that are not used to compute b_eff: p16 worst-cyc-1dim : 19.288 19.982 17.588 -> 19.982 -> 1278.822 MByte/s p17 best bi-section : 51.463 70.863 48.843 -> 70.863 -> 4535.211 MByte/s p18 worst bi-section : 18.041 20.779 16.803 -> 20.779 -> 1329.879 MByte/s p19 acyclic-1dim-all : 53.704 66.533 45.558 -> 66.533 -> 4258.127 MByte/s p20 acyclic-2dim-all : 42.968 50.379 37.278 -> 50.379 -> 3224.287 MByte/s p21 acyclic-3dim-all : 36.536 39.502 31.315 -> 39.502 -> 2528.135 MByte/s p22 cyclic-1dim-all : 53.692 70.396 45.225 -> 70.396 -> 4505.354 MByte/s p23 cyclic-2dim-all : 43.932 52.686 40.213 -> 52.686 -> 3371.879 MByte/s p24 cyclic-3dim-all : 37.180 41.849 32.430 -> 41.849 -> 2678.366 MByte/s log_avg of all rings : 57.045 71.135 47.074 || 71.354 -> 4566.665 MByte/s log_avg of all random : 30.683 32.164 27.328 || 32.169 -> 2058.846 MByte/s log_avg(ring,random) : 41.837 47.833 35.867 ||( 47.911 -> 3066.278)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-32*2fix : 72.508 72.486 72.271 -> 72.508 -> 4640.539 MByte/s p01 ring-16*4fix : 74.832 75.418 73.468 -> 75.418 -> 4826.757 MByte/s p02 ring-8*8fix : 67.099 66.737 66.608 -> 67.099 -> 4294.330 MByte/s p03 ring-4*16fix : 64.110 62.615 63.623 -> 64.110 -> 4103.037 MByte/s p04 ring-2*32fix : 77.571 77.634 77.721 -> 77.721 -> 4974.136 MByte/s p05 ring-1*64fix : 69.273 69.595 69.551 -> 69.595 -> 4454.086 MByte/s p06 random-cyc-1dim : 30.711 31.175 30.550 -> 31.175 -> 1995.206 MByte/s p07 random-cyc-1dim : 29.389 28.950 29.012 -> 29.389 -> 1880.891 MByte/s p08 random-cyc-1dim : 35.570 35.645 35.632 -> 35.645 -> 2281.257 MByte/s p09 random-cyc-1dim : 25.843 26.054 25.906 -> 26.054 -> 1667.475 MByte/s p10 random-cyc-1dim : 37.411 37.360 37.989 -> 37.989 -> 2431.280 MByte/s p11 random-cyc-1dim : 31.751 31.937 31.670 -> 31.937 -> 2043.978 MByte/s p12 random-cyc-1dim : 41.305 41.594 41.548 -> 41.594 -> 2662.038 MByte/s p13 random-cyc-1dim : 29.908 30.599 30.350 -> 30.599 -> 1958.312 MByte/s p14 random-cyc-1dim : 36.903 36.410 37.304 -> 37.304 -> 2387.444 MByte/s p15 random-cyc-1dim : 37.848 37.340 37.467 -> 37.848 -> 2422.258 MByte/s -- additional patterns that are not used to compute b_eff: p16 worst-cyc-1dim : 20.472 20.695 20.517 -> 20.695 -> 1324.480 MByte/s p17 best bi-section : 70.605 70.165 69.786 -> 70.605 -> 4518.746 MByte/s p18 worst bi-section : 20.708 20.709 20.695 -> 20.709 -> 1325.360 MByte/s p19 acyclic-1dim-all : 65.807 66.322 66.210 -> 66.322 -> 4244.628 MByte/s p20 acyclic-2dim-all : 52.282 52.474 51.937 -> 52.474 -> 3358.309 MByte/s p21 acyclic-3dim-all : 43.649 43.430 43.380 -> 43.649 -> 2793.509 MByte/s p22 cyclic-1dim-all : 68.704 69.172 69.459 -> 69.459 -> 4445.385 MByte/s p23 cyclic-2dim-all : 52.493 52.499 52.576 -> 52.576 -> 3364.861 MByte/s p24 cyclic-3dim-all : 44.280 44.136 44.242 -> 44.280 -> 2833.905 MByte/s log_avg of all rings : 70.751 70.561 70.389 || 70.920 -> 4538.895 MByte/s log_avg of all random : 33.347 33.405 33.414 || 33.638 -> 2152.854 MByte/s log_avg(ring,random) : 48.573 48.550 48.497 ||( 48.843 -> 3125.952)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-32*2fix p00 method 0 : 0.055 0.869 10.347 51.516 179.399 207.744 -> 72.164 -> 4618.495 MByte/s p00 method 1 : 0.018 0.304 4.660 55.986 180.592 208.034 -> 70.847 -> 4534.205 MByte/s p00 method 2 : 0.025 0.392 5.336 26.140 167.023 161.997 -> 60.632 -> 3880.464 MByte/s p01 ring-16*4fix p01 method 0 : 0.054 0.850 9.808 49.713 155.180 153.630 -> 59.560 -> 3811.810 MByte/s p01 method 1 : 0.030 0.516 7.750 79.899 185.140 193.977 -> 75.417 -> 4826.718 MByte/s p01 method 2 : 0.022 0.348 4.532 29.423 119.095 111.529 -> 45.450 -> 2908.828 MByte/s p02 ring-8*8fix p02 method 0 : 0.052 0.827 9.367 48.079 134.209 139.565 -> 53.474 -> 3422.311 MByte/s p02 method 1 : 0.030 0.516 7.712 76.365 152.510 186.591 -> 67.317 -> 4308.262 MByte/s p02 method 2 : 0.022 0.340 4.471 29.822 114.318 117.848 -> 44.340 -> 2837.755 MByte/s p03 ring-4*16fix p03 method 0 : 0.049 0.783 8.611 40.315 130.675 130.152 -> 51.115 -> 3271.386 MByte/s p03 method 1 : 0.030 0.513 7.649 76.621 140.640 170.961 -> 64.777 -> 4145.708 MByte/s p03 method 2 : 0.022 0.343 4.458 29.732 109.446 107.380 -> 42.772 -> 2737.392 MByte/s p04 ring-2*32fix p04 method 0 : 0.048 0.793 8.517 39.746 142.067 132.096 -> 54.760 -> 3504.621 MByte/s p04 method 1 : 0.030 0.516 7.713 80.484 193.232 211.855 -> 78.580 -> 5029.128 MByte/s p04 method 2 : 0.022 0.345 4.506 29.939 126.236 106.509 -> 46.037 -> 2946.385 MByte/s p05 ring-1*64fix p05 method 0 : 0.048 0.769 8.296 38.803 139.300 130.445 -> 53.563 -> 3428.027 MByte/s p05 method 1 : 0.029 0.510 7.639 80.425 166.635 183.225 -> 70.773 -> 4529.489 MByte/s p05 method 2 : 0.022 0.344 4.467 29.839 122.769 95.208 -> 45.224 -> 2894.321 MByte/s p06 random-cyc-1dim p06 method 0 : 0.044 0.719 7.754 35.437 63.417 61.209 -> 28.586 -> 1829.512 MByte/s p06 method 1 : 0.029 0.510 7.650 56.316 60.017 40.215 -> 28.537 -> 1826.391 MByte/s p06 method 2 : 0.021 0.334 4.308 29.052 60.066 56.495 -> 25.506 -> 1632.362 MByte/s p07 random-cyc-1dim p07 method 0 : 0.044 0.697 7.673 35.452 61.125 45.574 -> 26.958 -> 1725.305 MByte/s p07 method 1 : 0.029 0.504 7.572 55.123 59.775 41.852 -> 28.440 -> 1820.155 MByte/s p07 method 2 : 0.021 0.332 4.287 28.973 55.147 40.702 -> 23.193 -> 1484.348 MByte/s p08 random-cyc-1dim p08 method 0 : 0.045 0.699 7.681 36.081 73.943 69.820 -> 32.480 -> 2078.723 MByte/s p08 method 1 : 0.029 0.508 7.625 65.045 71.232 62.834 -> 34.341 -> 2197.816 MByte/s p08 method 2 : 0.021 0.335 4.348 28.943 70.130 58.076 -> 28.189 -> 1804.080 MByte/s p09 random-cyc-1dim p09 method 0 : 0.043 0.687 7.711 34.138 51.440 43.110 -> 23.858 -> 1526.922 MByte/s p09 method 1 : 0.029 0.501 7.503 49.613 52.192 41.708 -> 25.474 -> 1630.312 MByte/s p09 method 2 : 0.021 0.336 4.307 28.284 51.706 45.441 -> 22.284 -> 1426.186 MByte/s p10 random-cyc-1dim p10 method 0 : 0.045 0.729 7.741 37.240 82.156 64.911 -> 34.097 -> 2182.206 MByte/s p10 method 1 : 0.030 0.509 7.628 66.799 78.063 65.635 -> 36.874 -> 2359.913 MByte/s p10 method 2 : 0.021 0.336 4.339 29.220 74.473 69.097 -> 30.249 -> 1935.942 MByte/s p11 random-cyc-1dim p11 method 0 : 0.045 0.693 7.762 36.412 66.525 56.560 -> 29.253 -> 1872.216 MByte/s p11 method 1 : 0.030 0.512 7.674 59.417 62.456 40.239 -> 29.500 -> 1887.985 MByte/s p11 method 2 : 0.021 0.335 4.328 28.877 62.878 48.618 -> 25.899 -> 1657.517 MByte/s p12 random-cyc-1dim p12 method 0 : 0.044 0.686 7.818 37.306 93.762 84.435 -> 37.790 -> 2418.578 MByte/s p12 method 1 : 0.030 0.510 7.661 67.528 89.853 76.765 -> 40.685 -> 2603.867 MByte/s p12 method 2 : 0.021 0.336 4.327 29.160 83.432 83.233 -> 33.129 -> 2120.276 MByte/s p13 random-cyc-1dim p13 method 0 : 0.044 0.698 7.744 35.805 63.325 48.953 -> 27.767 -> 1777.093 MByte/s p13 method 1 : 0.029 0.505 7.561 56.424 60.857 52.266 -> 29.703 -> 1901.005 MByte/s p13 method 2 : 0.021 0.333 4.309 28.827 60.927 55.109 -> 25.400 -> 1625.608 MByte/s p14 random-cyc-1dim p14 method 0 : 0.044 0.712 7.819 36.894 81.811 83.894 -> 33.984 -> 2174.993 MByte/s p14 method 1 : 0.029 0.505 7.594 64.181 74.976 47.494 -> 34.184 -> 2187.785 MByte/s p14 method 2 : 0.021 0.333 4.327 28.993 73.681 74.948 -> 30.548 -> 1955.043 MByte/s p15 random-cyc-1dim p15 method 0 : 0.045 0.720 7.988 37.393 87.989 71.734 -> 34.817 -> 2228.264 MByte/s p15 method 1 : 0.030 0.509 7.611 62.750 79.939 66.410 -> 37.141 -> 2377.018 MByte/s p15 method 2 : 0.021 0.337 4.369 29.254 74.373 74.408 -> 31.045 -> 1986.906 MByte/s -- additional patterns that are not used to compute b_eff: p16 worst-cyc-1dim p16 method 0 : 0.043 0.674 7.639 30.652 40.667 29.134 -> 19.288 -> 1234.462 MByte/s p16 method 1 : 0.029 0.505 7.552 38.867 40.128 28.142 -> 19.982 -> 1278.822 MByte/s p16 method 2 : 0.021 0.332 4.229 26.834 37.340 29.353 -> 17.588 -> 1125.614 MByte/s p17 best bi-section p17 method 0 : 0.033 0.497 5.518 28.264 130.947 161.482 -> 51.463 -> 3293.631 MByte/s p17 method 1 : 0.018 0.302 4.613 55.586 179.780 208.103 -> 70.863 -> 4535.211 MByte/s p17 method 2 : 0.015 0.241 3.347 24.227 124.274 161.316 -> 48.843 -> 3125.925 MByte/s p18 worst bi-section p18 method 0 : 0.024 0.368 4.441 25.074 37.441 38.333 -> 18.041 -> 1154.638 MByte/s p18 method 1 : 0.018 0.296 4.546 33.468 39.230 56.095 -> 20.779 -> 1329.879 MByte/s p18 method 2 : 0.015 0.239 3.312 22.187 36.628 35.516 -> 16.803 -> 1075.404 MByte/s p19 acyclic-1dim-all p19 method 0 : 0.047 0.762 8.284 38.384 139.967 129.885 -> 53.704 -> 3437.049 MByte/s p19 method 1 : 0.030 0.509 7.647 77.639 165.929 144.662 -> 66.533 -> 4258.127 MByte/s p19 method 2 : 0.021 0.339 4.405 29.489 122.334 112.116 -> 45.558 -> 2915.723 MByte/s p20 acyclic-2dim-all p20 method 0 : 0.036 0.574 6.232 28.699 109.365 116.468 -> 42.968 -> 2749.955 MByte/s p20 method 1 : 0.041 0.724 10.398 81.457 114.420 87.272 -> 50.379 -> 3224.287 MByte/s p20 method 2 : 0.019 0.302 3.913 26.347 97.655 84.317 -> 37.278 -> 2385.766 MByte/s p21 acyclic-3dim-all p21 method 0 : 0.030 0.483 5.288 24.330 93.986 93.973 -> 36.536 -> 2338.321 MByte/s p21 method 1 : 0.043 0.763 10.711 73.127 83.982 75.917 -> 39.502 -> 2528.135 MByte/s p21 method 2 : 0.016 0.257 3.312 22.153 84.400 68.521 -> 31.315 -> 2004.163 MByte/s p22 cyclic-1dim-all p22 method 0 : 0.047 0.753 8.566 39.504 140.563 131.850 -> 53.692 -> 3436.290 MByte/s p22 method 1 : 0.029 0.495 7.438 78.258 167.925 183.468 -> 70.396 -> 4505.354 MByte/s p22 method 2 : 0.022 0.346 4.491 29.708 124.045 102.611 -> 45.225 -> 2894.392 MByte/s p23 cyclic-2dim-all p23 method 0 : 0.040 0.634 6.870 32.801 111.451 105.893 -> 43.932 -> 2811.657 MByte/s p23 method 1 : 0.043 0.770 11.110 82.028 115.529 105.681 -> 52.686 -> 3371.879 MByte/s p23 method 2 : 0.022 0.343 4.346 29.697 105.389 90.877 -> 40.213 -> 2573.664 MByte/s p24 cyclic-3dim-all p24 method 0 : 0.038 0.596 6.521 30.954 91.615 94.180 -> 37.180 -> 2379.531 MByte/s p24 method 1 : 0.053 0.938 13.161 71.884 83.066 80.502 -> 41.849 -> 2678.366 MByte/s p24 method 2 : 0.021 0.333 4.307 28.554 81.270 68.823 -> 32.430 -> 2075.519 MByte/s log_avg of all rings - ring, method 0 : 0.051 0.814 9.128 44.395 145.942 146.755 || 57.045 -> 3650.856 MByte/s - ring, method 1 : 0.027 0.471 7.076 74.387 168.749 191.918 || 71.135 -> 4552.668 MByte/s - ring, method 2 : 0.022 0.352 4.618 29.116 125.230 115.045 || 47.074 -> 3012.717 MByte/s log_avg of all random - random, method 0 : 0.044 0.704 7.769 36.202 71.402 61.442 || 30.683 -> 1963.704 MByte/s - random, method 1 : 0.029 0.507 7.608 60.054 68.055 52.102 || 32.164 -> 2058.495 MByte/s - random, method 2 : 0.021 0.335 4.325 28.957 65.993 59.118 || 27.328 -> 1748.991 MByte/s log_avg(ring,random) - average, method 0 : 0.048 0.757 8.421 40.090 102.081 94.958 || 41.837 -> 2677.537 MByte/s - average, method 1 : 0.028 0.489 7.337 66.837 107.165 99.997 || 47.833 -> 3061.314 MByte/s - average, method 2 : 0.022 0.343 4.469 29.036 90.909 82.470 || 35.867 -> 2295.477 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 3.041 0.048 0.051 0.044 0.048 0.028 0.022 2 6.058 0.095 0.102 0.088 0.095 0.057 0.043 4 12.166 0.190 0.204 0.177 0.190 0.125 0.087 8 24.289 0.380 0.406 0.355 0.380 0.249 0.173 16 48.447 0.757 0.814 0.704 0.757 0.489 0.343 32 73.589 1.150 1.272 1.039 1.150 0.963 0.598 64 134.608 2.103 2.300 1.923 2.099 1.854 1.134 128 294.241 4.598 5.068 4.171 4.598 3.834 2.359 256 538.929 8.421 9.128 7.769 8.421 7.337 4.469 512 980.987 15.328 16.195 14.508 14.955 14.040 8.130 1024 1749.533 27.336 27.751 26.928 22.891 26.265 13.987 2048 2918.456 45.601 46.941 44.299 31.312 44.881 21.604 4096 4277.593 66.837 74.387 60.054 40.090 66.837 29.036 8192 5458.027 85.282 107.410 67.712 62.262 85.282 58.668 16384 6262.513 97.852 136.911 69.936 81.419 97.852 75.538 32768 6774.963 105.859 159.022 70.469 94.743 105.658 86.016 65536 7030.241 109.848 168.749 71.505 102.081 107.165 90.909 131072 7173.294 112.083 175.161 71.720 104.709 106.273 92.508 262144 7250.792 113.294 185.616 69.150 103.082 107.635 91.693 524288 7252.521 113.321 191.582 67.029 100.636 104.878 87.511 1048576 7044.317 110.067 191.918 63.125 94.958 99.997 82.470 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-32*2fix : 0.055 0.869 10.347 55.986 180.592 208.034 -> 72.676 -> 4651.252 MByte/s p01 ring-16*4fix : 0.054 0.850 9.808 79.899 185.140 193.977 -> 75.775 -> 4849.574 MByte/s p02 ring-8*8fix : 0.052 0.827 9.367 76.365 152.510 186.591 -> 67.627 -> 4328.158 MByte/s p03 ring-4*16fix : 0.049 0.783 8.611 76.621 140.640 170.961 -> 64.912 -> 4154.347 MByte/s p04 ring-2*32fix : 0.048 0.793 8.517 80.484 193.232 211.855 -> 78.757 -> 5040.427 MByte/s p05 ring-1*64fix : 0.048 0.769 8.296 80.425 166.635 183.225 -> 70.873 -> 4535.888 MByte/s p06 random-cyc-1dim : 0.044 0.719 7.754 56.316 63.417 61.209 -> 31.378 -> 2008.200 MByte/s p07 random-cyc-1dim : 0.044 0.697 7.673 55.123 61.125 45.574 -> 29.554 -> 1891.469 MByte/s p08 random-cyc-1dim : 0.045 0.699 7.681 65.045 73.943 69.820 -> 36.208 -> 2317.332 MByte/s p09 random-cyc-1dim : 0.043 0.687 7.711 49.613 52.192 45.441 -> 26.315 -> 1684.157 MByte/s p10 random-cyc-1dim : 0.045 0.729 7.741 66.799 82.156 69.097 -> 38.312 -> 2451.972 MByte/s p11 random-cyc-1dim : 0.045 0.693 7.762 59.417 66.525 56.560 -> 32.236 -> 2063.102 MByte/s p12 random-cyc-1dim : 0.044 0.686 7.818 67.528 93.762 84.435 -> 42.251 -> 2704.064 MByte/s p13 random-cyc-1dim : 0.044 0.698 7.744 56.424 63.325 55.109 -> 30.860 -> 1975.014 MByte/s p14 random-cyc-1dim : 0.044 0.712 7.819 64.181 81.811 83.894 -> 37.798 -> 2419.093 MByte/s p15 random-cyc-1dim : 0.045 0.720 7.988 62.750 87.989 74.408 -> 38.487 -> 2463.190 MByte/s -- additional patterns that are not used to compute b_eff: p16 worst-cyc-1dim : 0.043 0.674 7.639 38.867 40.667 29.353 -> 20.913 -> 1338.406 MByte/s p17 best bi-section : 0.033 0.497 5.518 55.586 179.780 208.103 -> 71.002 -> 4544.133 MByte/s p18 worst bi-section : 0.024 0.368 4.546 33.468 39.230 56.095 -> 20.835 -> 1333.457 MByte/s p19 acyclic-1dim-all : 0.047 0.762 8.284 77.639 165.929 144.662 -> 66.628 -> 4264.162 MByte/s p20 acyclic-2dim-all : 0.041 0.724 10.398 81.457 114.420 116.468 -> 52.740 -> 3375.364 MByte/s p21 acyclic-3dim-all : 0.043 0.763 10.711 73.127 93.986 93.973 -> 44.222 -> 2830.195 MByte/s p22 cyclic-1dim-all : 0.047 0.753 8.566 78.258 167.925 183.468 -> 70.530 -> 4513.903 MByte/s p23 cyclic-2dim-all : 0.043 0.770 11.110 82.028 115.529 105.893 -> 53.111 -> 3399.080 MByte/s p24 cyclic-3dim-all : 0.053 0.938 13.161 71.884 91.615 94.180 -> 44.560 -> 2851.831 MByte/s log_avg of all rings : 0.051 0.814 9.128 74.387 168.749 191.918 || 71.618 -> 4583.545 MByte/s log_avg of all random : 0.044 0.704 7.769 60.054 71.505 63.125 || 34.009 -> 2176.582 MByte/s log_avg(ring,random) : 0.048 0.757 8.421 66.837 109.848 110.067 || 49.352 -> 3158.554 MByte/s SECTION-BY-PATTERN-MSGLNG-END SECTION-BEFF-BEGIN The effective bandwidth is b_eff = 3158.554 MByte/s on 64 processes ( = 49.352 MByte/s * 64 processes) system parameters : 64 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 = 3158.554 MB/s = 49.352 * 64 PEs with 128 MB/PE on sn6715 hwwt3e 2.0.4.71 unicosmk CRAY T3E