b_eff = 10056.033 MB/s = 39.281 * 256 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 = 256 1-dim-paterns: size = 256 2-dim-paterns: size = 16 * 16 3-dim-paterns: size = 8 * 8 * 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 = 218.111 sec sum of max elapsed time per entries above = 225.437 sec difference = -7.326 sec = 3.4% The difference is less than 5 % SECTION-LOOP-END SECTION-PATTERN-BEGIN Pattern parameters: p00 ring-128*2fix => 1 sendrecv_calls with 256 messages, i.e. msgs/used node, all nodes are used p01 ring-64*4fix => 2 sendrecv_calls with 512 messages, i.e. msgs/used node, all nodes are used p02 ring-32*8fix => 2 sendrecv_calls with 512 messages, i.e. msgs/used node, all nodes are used p03 ring-4*64fix => 2 sendrecv_calls with 512 messages, i.e. msgs/used node, all nodes are used p04 ring-2*128fix => 2 sendrecv_calls with 512 messages, i.e. msgs/used node, all nodes are used p05 ring-1*256fix => 2 sendrecv_calls with 512 messages, i.e. msgs/used node, all nodes are used p06 random-cyc-1dim => 2 sendrecv_calls with 512 messages, i.e. msgs/used node, all nodes are used p07 random-cyc-1dim => 2 sendrecv_calls with 512 messages, i.e. msgs/used node, all nodes are used p08 random-cyc-1dim => 2 sendrecv_calls with 512 messages, i.e. msgs/used node, all nodes are used p09 random-cyc-1dim => 2 sendrecv_calls with 512 messages, i.e. msgs/used node, all nodes are used p10 random-cyc-1dim => 2 sendrecv_calls with 512 messages, i.e. msgs/used node, all nodes are used p11 random-cyc-1dim => 2 sendrecv_calls with 512 messages, i.e. msgs/used node, all nodes are used p12 random-cyc-1dim => 2 sendrecv_calls with 512 messages, i.e. msgs/used node, all nodes are used p13 random-cyc-1dim => 2 sendrecv_calls with 512 messages, i.e. msgs/used node, all nodes are used p14 random-cyc-1dim => 2 sendrecv_calls with 512 messages, i.e. msgs/used node, all nodes are used p15 random-cyc-1dim => 2 sendrecv_calls with 512 messages, i.e. msgs/used node, all nodes are used p16 worst-cyc-1dim => 2 sendrecv_calls with 512 messages, i.e. msgs/used node, all nodes are used p17 best bi-section => 2 sendrecv_calls with 256 messages, i.e. msgs/used node, all nodes are used p18 worst bi-section => 2 sendrecv_calls with 256 messages, i.e. msgs/used node, all nodes are used p19 acyclic-1dim-all => 2 sendrecv_calls with 510 messages, i.e. msgs/used node, all nodes are used p20 acyclic-2dim-all => 4 sendrecv_calls with 960 messages, i.e. msgs/used node, all nodes are used p21 acyclic-3dim-all => 6 sendrecv_calls with 1280 messages, i.e. msgs/used node, all nodes are used p22 cyclic-1dim-all => 2 sendrecv_calls with 512 messages, i.e. msgs/used node, all nodes are used p23 cyclic-2dim-all => 4 sendrecv_calls with 1024 messages, i.e. msgs/used node, all nodes are used p24 cyclic-3dim-all => 6 sendrecv_calls with 1536 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-128*2fix : 71.340 54.893 58.272 -> 71.340 -> 18263.098 MByte/s p01 ring-64*4fix : 57.448 59.519 44.262 -> 59.519 -> 15236.941 MByte/s p02 ring-32*8fix : 51.239 54.956 42.914 -> 54.956 -> 14068.738 MByte/s p03 ring-4*64fix : 52.975 57.095 43.692 -> 57.095 -> 14616.213 MByte/s p04 ring-2*128fix : 52.587 56.446 43.752 -> 56.446 -> 14450.200 MByte/s p05 ring-1*256fix : 53.273 59.525 44.159 -> 59.525 -> 15238.359 MByte/s p06 random-cyc-1dim : 21.337 17.879 18.188 -> 21.337 -> 5462.155 MByte/s p07 random-cyc-1dim : 24.620 20.441 22.090 -> 24.620 -> 6302.760 MByte/s p08 random-cyc-1dim : 25.120 20.527 22.019 -> 25.120 -> 6430.824 MByte/s p09 random-cyc-1dim : 26.297 21.966 22.804 -> 26.297 -> 6732.080 MByte/s p10 random-cyc-1dim : 27.032 24.016 24.050 -> 27.032 -> 6920.165 MByte/s p11 random-cyc-1dim : 22.219 17.160 18.370 -> 22.219 -> 5688.044 MByte/s p12 random-cyc-1dim : 22.650 18.694 19.524 -> 22.650 -> 5798.490 MByte/s p13 random-cyc-1dim : 22.912 19.333 19.943 -> 22.912 -> 5865.363 MByte/s p14 random-cyc-1dim : 27.585 23.740 23.712 -> 27.585 -> 7061.784 MByte/s p15 random-cyc-1dim : 26.666 22.177 23.122 -> 26.666 -> 6826.450 MByte/s -- additional patterns that are not used to compute b_eff: p16 worst-cyc-1dim : 14.925 13.372 13.490 -> 14.925 -> 3820.741 MByte/s p17 best bi-section : 49.989 54.825 48.371 -> 54.825 -> 14035.269 MByte/s p18 worst bi-section : 14.111 14.816 12.911 -> 14.816 -> 3792.961 MByte/s p19 acyclic-1dim-all : 53.160 53.887 44.305 -> 53.887 -> 13795.088 MByte/s p20 acyclic-2dim-all : 37.543 37.884 32.011 -> 37.884 -> 9698.393 MByte/s p21 acyclic-3dim-all : 43.168 43.360 34.998 -> 43.360 -> 11100.252 MByte/s p22 cyclic-1dim-all : 53.374 59.824 44.643 -> 59.824 -> 15314.936 MByte/s p23 cyclic-2dim-all : 38.319 37.917 32.601 -> 38.319 -> 9809.757 MByte/s p24 cyclic-3dim-all : 47.335 44.616 38.869 -> 47.335 -> 12117.767 MByte/s log_avg of all rings : 56.097 57.041 45.893 || 59.588 -> 15254.421 MByte/s log_avg of all random : 24.551 20.471 21.278 || 24.551 -> 6285.082 MByte/s log_avg(ring,random) : 37.111 34.171 31.249 ||( 38.248 -> 9791.592)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-128*2fix : 70.674 70.291 68.986 -> 70.674 -> 18092.493 MByte/s p01 ring-64*4fix : 63.113 64.062 63.348 -> 64.062 -> 16399.811 MByte/s p02 ring-32*8fix : 57.412 58.468 57.883 -> 58.468 -> 14967.805 MByte/s p03 ring-4*64fix : 57.866 58.138 58.878 -> 58.878 -> 15072.647 MByte/s p04 ring-2*128fix : 57.394 58.261 57.744 -> 58.261 -> 14914.797 MByte/s p05 ring-1*256fix : 61.588 61.002 61.685 -> 61.685 -> 15791.303 MByte/s p06 random-cyc-1dim : 21.048 21.077 20.886 -> 21.077 -> 5395.690 MByte/s p07 random-cyc-1dim : 24.587 24.367 23.920 -> 24.587 -> 6294.310 MByte/s p08 random-cyc-1dim : 24.274 24.884 24.638 -> 24.884 -> 6370.270 MByte/s p09 random-cyc-1dim : 26.099 26.010 25.732 -> 26.099 -> 6681.230 MByte/s p10 random-cyc-1dim : 26.362 26.572 26.595 -> 26.595 -> 6808.320 MByte/s p11 random-cyc-1dim : 21.479 21.685 22.077 -> 22.077 -> 5651.791 MByte/s p12 random-cyc-1dim : 22.167 22.359 22.309 -> 22.359 -> 5723.903 MByte/s p13 random-cyc-1dim : 22.184 22.992 22.294 -> 22.992 -> 5885.948 MByte/s p14 random-cyc-1dim : 27.032 26.829 27.034 -> 27.034 -> 6920.744 MByte/s p15 random-cyc-1dim : 26.413 25.913 26.438 -> 26.438 -> 6768.055 MByte/s -- additional patterns that are not used to compute b_eff: p16 worst-cyc-1dim : 14.723 14.922 14.806 -> 14.922 -> 3820.051 MByte/s p17 best bi-section : 57.610 56.631 55.800 -> 57.610 -> 14748.187 MByte/s p18 worst bi-section : 16.125 16.288 16.173 -> 16.288 -> 4169.736 MByte/s p19 acyclic-1dim-all : 55.841 55.841 55.289 -> 55.841 -> 14295.261 MByte/s p20 acyclic-2dim-all : 40.098 39.483 39.811 -> 40.098 -> 10265.108 MByte/s p21 acyclic-3dim-all : 47.576 47.633 47.656 -> 47.656 -> 12199.949 MByte/s p22 cyclic-1dim-all : 61.827 61.804 61.716 -> 61.827 -> 15827.723 MByte/s p23 cyclic-2dim-all : 40.237 39.814 39.983 -> 40.237 -> 10300.714 MByte/s p24 cyclic-3dim-all : 50.900 51.059 50.316 -> 51.059 -> 13070.998 MByte/s log_avg of all rings : 61.169 61.555 61.298 || 61.856 -> 15835.072 MByte/s log_avg of all random : 24.066 24.185 24.100 || 24.327 -> 6227.820 MByte/s log_avg(ring,random) : 38.368 38.584 38.436 ||( 38.792 -> 9930.659)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-128*2fix p00 method 0 : 0.056 0.868 9.813 51.383 179.296 199.711 -> 71.340 -> 18263.098 MByte/s p00 method 1 : 0.005 0.087 1.370 20.559 140.266 199.025 -> 54.893 -> 14052.505 MByte/s p00 method 2 : 0.025 0.396 5.230 25.555 163.923 159.708 -> 58.272 -> 14917.599 MByte/s p01 ring-64*4fix p01 method 0 : 0.054 0.841 8.830 49.004 146.994 153.540 -> 57.448 -> 14706.601 MByte/s p01 method 1 : 0.010 0.164 2.571 35.555 154.594 192.168 -> 59.519 -> 15236.941 MByte/s p01 method 2 : 0.021 0.340 4.483 29.017 116.888 112.713 -> 44.262 -> 11330.964 MByte/s p02 ring-32*8fix p02 method 0 : 0.052 0.750 9.480 46.781 133.444 107.787 -> 51.239 -> 13117.303 MByte/s p02 method 1 : 0.010 0.165 2.572 35.301 139.313 182.304 -> 54.956 -> 14068.738 MByte/s p02 method 2 : 0.021 0.328 4.343 29.742 112.465 107.119 -> 42.914 -> 10986.067 MByte/s p03 ring-4*64fix p03 method 0 : 0.047 0.698 8.220 37.521 137.901 129.193 -> 52.975 -> 13561.573 MByte/s p03 method 1 : 0.010 0.164 2.569 35.465 143.921 182.958 -> 57.095 -> 14616.213 MByte/s p03 method 2 : 0.022 0.321 4.488 29.670 121.817 95.502 -> 43.692 -> 11185.038 MByte/s p04 ring-2*128fix p04 method 0 : 0.046 0.749 8.208 36.723 138.581 129.875 -> 52.587 -> 13462.149 MByte/s p04 method 1 : 0.010 0.164 2.569 35.531 139.967 182.565 -> 56.446 -> 14450.200 MByte/s p04 method 2 : 0.021 0.341 4.283 29.435 120.627 97.304 -> 43.752 -> 11200.424 MByte/s p05 ring-1*256fix p05 method 0 : 0.047 0.728 8.216 38.171 141.089 130.389 -> 53.273 -> 13637.763 MByte/s p05 method 1 : 0.010 0.165 2.568 35.429 153.479 199.063 -> 59.525 -> 15238.359 MByte/s p05 method 2 : 0.021 0.344 4.468 29.671 120.824 98.898 -> 44.159 -> 11304.696 MByte/s p06 random-cyc-1dim p06 method 0 : 0.041 0.665 7.308 33.747 46.980 27.516 -> 21.337 -> 5462.155 MByte/s p06 method 1 : 0.010 0.163 2.539 33.493 41.386 23.067 -> 17.879 -> 4576.903 MByte/s p06 method 2 : 0.020 0.322 4.021 27.145 42.568 26.192 -> 18.188 -> 4656.034 MByte/s p07 random-cyc-1dim p07 method 0 : 0.039 0.660 7.338 34.303 55.917 42.245 -> 24.620 -> 6302.760 MByte/s p07 method 1 : 0.010 0.164 2.542 33.604 51.565 27.505 -> 20.441 -> 5232.867 MByte/s p07 method 2 : 0.020 0.322 4.027 28.400 52.440 41.051 -> 22.090 -> 5655.100 MByte/s p08 random-cyc-1dim p08 method 0 : 0.041 0.614 7.099 35.035 57.386 44.064 -> 25.120 -> 6430.824 MByte/s p08 method 1 : 0.010 0.163 2.550 34.003 50.686 25.646 -> 20.527 -> 5255.018 MByte/s p08 method 2 : 0.020 0.323 4.042 28.485 52.153 34.073 -> 22.019 -> 5636.773 MByte/s p09 random-cyc-1dim p09 method 0 : 0.042 0.659 7.095 35.065 60.791 46.355 -> 26.297 -> 6732.080 MByte/s p09 method 1 : 0.010 0.164 2.561 33.875 54.886 29.858 -> 21.966 -> 5623.423 MByte/s p09 method 2 : 0.021 0.315 4.074 28.006 54.658 38.449 -> 22.804 -> 5837.923 MByte/s p10 random-cyc-1dim p10 method 0 : 0.042 0.670 7.426 34.569 60.612 51.785 -> 27.032 -> 6920.165 MByte/s p10 method 1 : 0.010 0.164 2.556 33.860 59.755 44.829 -> 24.016 -> 6148.212 MByte/s p10 method 2 : 0.021 0.314 4.162 27.734 58.236 47.810 -> 24.050 -> 6156.733 MByte/s p11 random-cyc-1dim p11 method 0 : 0.042 0.657 7.350 33.151 49.036 35.852 -> 22.219 -> 5688.044 MByte/s p11 method 1 : 0.010 0.163 2.553 33.358 39.007 24.007 -> 17.160 -> 4393.066 MByte/s p11 method 2 : 0.020 0.321 4.052 27.108 40.352 32.934 -> 18.370 -> 4702.669 MByte/s p12 random-cyc-1dim p12 method 0 : 0.041 0.641 7.319 34.136 51.475 33.837 -> 22.650 -> 5798.490 MByte/s p12 method 1 : 0.010 0.164 2.559 33.954 44.731 23.714 -> 18.694 -> 4785.548 MByte/s p12 method 2 : 0.020 0.317 4.163 28.163 43.861 30.908 -> 19.524 -> 4998.018 MByte/s p13 random-cyc-1dim p13 method 0 : 0.042 0.624 7.031 33.852 49.806 43.066 -> 22.912 -> 5865.363 MByte/s p13 method 1 : 0.010 0.163 2.548 33.718 46.271 25.758 -> 19.333 -> 4949.307 MByte/s p13 method 2 : 0.020 0.308 4.012 27.901 45.320 31.434 -> 19.943 -> 5105.342 MByte/s p14 random-cyc-1dim p14 method 0 : 0.042 0.651 7.418 34.863 64.025 51.787 -> 27.585 -> 7061.784 MByte/s p14 method 1 : 0.010 0.163 2.556 33.738 60.031 43.304 -> 23.740 -> 6077.372 MByte/s p14 method 2 : 0.020 0.322 4.149 27.838 57.133 45.413 -> 23.712 -> 6070.274 MByte/s p15 random-cyc-1dim p15 method 0 : 0.042 0.660 7.180 34.431 62.515 46.656 -> 26.666 -> 6826.450 MByte/s p15 method 1 : 0.010 0.164 2.557 34.033 55.238 31.768 -> 22.177 -> 5677.374 MByte/s p15 method 2 : 0.019 0.323 3.979 28.452 53.815 42.472 -> 23.122 -> 5919.326 MByte/s -- additional patterns that are not used to compute b_eff: p16 worst-cyc-1dim p16 method 0 : 0.041 0.597 7.302 26.191 28.781 23.671 -> 14.925 -> 3820.741 MByte/s p16 method 1 : 0.010 0.163 2.549 27.223 28.530 21.886 -> 13.372 -> 3423.305 MByte/s p16 method 2 : 0.019 0.310 3.998 24.465 27.998 23.134 -> 13.490 -> 3453.420 MByte/s p17 best bi-section p17 method 0 : 0.032 0.494 5.490 26.837 126.257 155.898 -> 49.989 -> 12797.310 MByte/s p17 method 1 : 0.005 0.086 1.361 20.450 139.251 204.062 -> 54.825 -> 14035.269 MByte/s p17 method 2 : 0.014 0.236 3.166 24.120 124.865 157.006 -> 48.371 -> 12382.955 MByte/s p18 worst bi-section p18 method 0 : 0.023 0.349 4.437 23.268 28.857 29.364 -> 14.111 -> 3612.326 MByte/s p18 method 1 : 0.005 0.086 1.363 18.535 30.886 48.868 -> 14.816 -> 3792.961 MByte/s p18 method 2 : 0.013 0.212 2.881 20.992 29.524 24.246 -> 12.911 -> 3305.192 MByte/s p19 acyclic-1dim-all p19 method 0 : 0.046 0.750 8.335 39.142 140.617 129.799 -> 53.160 -> 13608.998 MByte/s p19 method 1 : 0.010 0.164 2.559 35.301 150.041 144.903 -> 53.887 -> 13795.088 MByte/s p19 method 2 : 0.021 0.340 4.272 29.685 121.113 97.844 -> 44.305 -> 11342.035 MByte/s p20 acyclic-2dim-all p20 method 0 : 0.036 0.575 6.121 29.140 95.286 94.436 -> 37.543 -> 9611.110 MByte/s p20 method 1 : 0.017 0.286 4.388 50.478 88.106 91.204 -> 37.884 -> 9698.393 MByte/s p20 method 2 : 0.020 0.318 4.063 27.431 81.737 67.144 -> 32.011 -> 8194.842 MByte/s p21 acyclic-3dim-all p21 method 0 : 0.032 0.504 5.568 25.599 110.898 116.329 -> 43.168 -> 11050.972 MByte/s p21 method 1 : 0.022 0.362 5.455 58.714 104.582 90.679 -> 43.360 -> 11100.252 MByte/s p21 method 2 : 0.018 0.275 3.649 24.467 93.753 76.505 -> 34.998 -> 8959.504 MByte/s p22 cyclic-1dim-all p22 method 0 : 0.046 0.746 8.065 39.374 138.347 130.208 -> 53.374 -> 13663.694 MByte/s p22 method 1 : 0.010 0.163 2.545 34.913 154.963 197.772 -> 59.824 -> 15314.936 MByte/s p22 method 2 : 0.022 0.341 4.135 29.585 122.475 97.527 -> 44.643 -> 11428.555 MByte/s p23 cyclic-2dim-all p23 method 0 : 0.038 0.582 6.118 31.001 95.910 93.687 -> 38.319 -> 9809.757 MByte/s p23 method 1 : 0.018 0.296 4.508 49.812 88.172 88.702 -> 37.917 -> 9706.728 MByte/s p23 method 2 : 0.020 0.324 4.220 28.042 83.745 65.266 -> 32.601 -> 8345.848 MByte/s p24 cyclic-3dim-all p24 method 0 : 0.037 0.586 6.572 31.336 123.859 123.504 -> 47.335 -> 12117.767 MByte/s p24 method 1 : 0.024 0.409 6.184 63.413 106.822 91.664 -> 44.616 -> 11421.755 MByte/s p24 method 2 : 0.021 0.339 4.230 29.060 103.707 80.589 -> 38.869 -> 9950.526 MByte/s log_avg of all rings - ring, method 0 : 0.050 0.770 8.771 42.856 145.486 139.059 || 56.097 -> 14360.949 MByte/s - ring, method 1 : 0.009 0.148 2.314 32.377 145.118 189.536 || 57.041 -> 14602.446 MByte/s - ring, method 2 : 0.022 0.344 4.539 28.807 125.061 110.006 || 45.893 -> 11748.700 MByte/s log_avg of all random - random, method 0 : 0.041 0.650 7.255 34.310 55.544 41.600 || 24.551 -> 6285.082 MByte/s - random, method 1 : 0.010 0.163 2.552 33.763 49.861 29.126 || 20.471 -> 5240.518 MByte/s - random, method 2 : 0.020 0.319 4.068 27.919 49.670 36.465 || 21.278 -> 5447.213 MByte/s log_avg(ring,random) - average, method 0 : 0.046 0.707 7.977 38.346 89.894 76.058 || 37.111 -> 9500.513 MByte/s - average, method 1 : 0.009 0.155 2.430 33.063 85.063 74.299 || 34.171 -> 8747.821 MByte/s - average, method 2 : 0.021 0.331 4.297 28.360 78.814 63.335 || 31.249 -> 7999.854 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 11.668 0.046 0.050 0.041 0.046 0.009 0.021 2 23.128 0.090 0.100 0.082 0.090 0.019 0.042 4 46.721 0.183 0.201 0.166 0.183 0.039 0.084 8 93.177 0.364 0.397 0.334 0.364 0.078 0.167 16 181.076 0.707 0.770 0.650 0.707 0.155 0.331 32 279.153 1.090 1.241 0.959 1.090 0.309 0.574 64 516.341 2.017 2.234 1.821 2.017 0.611 1.097 128 1121.837 4.382 4.923 3.900 4.382 1.234 2.283 256 2042.175 7.977 8.771 7.255 7.977 2.430 4.297 512 3620.366 14.142 15.504 12.900 14.142 4.786 7.851 1024 5626.391 21.978 24.611 19.627 21.978 9.384 13.636 2048 7709.595 30.116 33.295 27.240 30.116 17.916 20.840 4096 9819.544 38.358 42.856 34.331 38.346 33.063 28.360 8192 15044.227 58.767 71.516 48.290 57.454 52.298 54.162 16384 18788.629 73.393 99.974 53.880 72.999 68.167 67.762 32768 21560.134 84.219 128.023 55.403 83.887 78.268 75.223 65536 23458.785 91.636 151.179 55.544 89.894 85.063 78.814 131072 24531.695 95.827 168.318 54.556 91.599 86.829 79.396 262144 24714.435 96.541 181.533 51.341 88.865 86.686 76.396 524288 24092.951 94.113 186.641 47.456 84.613 82.132 71.639 1048576 22738.164 88.821 189.645 41.600 76.058 74.299 63.335 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-128*2fix : 0.056 0.868 9.813 51.383 179.296 199.711 -> 71.340 -> 18263.098 MByte/s p01 ring-64*4fix : 0.054 0.841 8.830 49.004 154.594 192.168 -> 64.936 -> 16623.589 MByte/s p02 ring-32*8fix : 0.052 0.750 9.480 46.781 139.313 182.304 -> 58.768 -> 15044.666 MByte/s p03 ring-4*64fix : 0.047 0.698 8.220 37.521 143.921 182.958 -> 59.668 -> 15275.076 MByte/s p04 ring-2*128fix : 0.046 0.749 8.208 36.723 139.967 182.565 -> 59.024 -> 15110.036 MByte/s p05 ring-1*256fix : 0.047 0.728 8.216 38.171 153.479 199.063 -> 62.312 -> 15951.763 MByte/s p06 random-cyc-1dim : 0.041 0.665 7.308 33.747 46.980 27.516 -> 21.454 -> 5492.257 MByte/s p07 random-cyc-1dim : 0.039 0.660 7.338 34.303 55.917 42.245 -> 24.773 -> 6341.940 MByte/s p08 random-cyc-1dim : 0.041 0.614 7.099 35.035 57.386 44.064 -> 25.329 -> 6484.280 MByte/s p09 random-cyc-1dim : 0.042 0.659 7.095 35.065 60.791 46.355 -> 26.435 -> 6767.476 MByte/s p10 random-cyc-1dim : 0.042 0.670 7.426 34.569 60.612 51.785 -> 27.068 -> 6929.438 MByte/s p11 random-cyc-1dim : 0.042 0.657 7.350 33.358 49.036 35.852 -> 22.286 -> 5705.173 MByte/s p12 random-cyc-1dim : 0.041 0.641 7.319 34.136 51.475 33.837 -> 22.771 -> 5829.363 MByte/s p13 random-cyc-1dim : 0.042 0.624 7.031 33.852 49.806 43.066 -> 23.084 -> 5909.595 MByte/s p14 random-cyc-1dim : 0.042 0.651 7.418 34.863 64.025 51.787 -> 27.673 -> 7084.326 MByte/s p15 random-cyc-1dim : 0.042 0.660 7.180 34.431 62.515 46.656 -> 26.834 -> 6869.571 MByte/s -- additional patterns that are not used to compute b_eff: p16 worst-cyc-1dim : 0.041 0.597 7.302 27.223 28.781 23.671 -> 15.048 -> 3852.309 MByte/s p17 best bi-section : 0.032 0.494 5.490 26.837 139.251 204.062 -> 58.178 -> 14893.552 MByte/s p18 worst bi-section : 0.023 0.349 4.437 23.268 30.886 48.868 -> 16.374 -> 4191.854 MByte/s p19 acyclic-1dim-all : 0.046 0.750 8.335 39.142 150.041 144.903 -> 56.801 -> 14540.994 MByte/s p20 acyclic-2dim-all : 0.036 0.575 6.121 50.478 95.286 94.436 -> 40.327 -> 10323.675 MByte/s p21 acyclic-3dim-all : 0.032 0.504 5.568 58.714 110.898 116.329 -> 47.999 -> 12287.686 MByte/s p22 cyclic-1dim-all : 0.046 0.746 8.065 39.374 154.963 197.772 -> 62.778 -> 16071.205 MByte/s p23 cyclic-2dim-all : 0.038 0.582 6.118 49.812 95.910 93.687 -> 40.496 -> 10367.048 MByte/s p24 cyclic-3dim-all : 0.037 0.586 6.572 63.413 123.859 123.504 -> 51.354 -> 13146.734 MByte/s log_avg of all rings : 0.050 0.770 8.771 42.856 151.179 189.645 || 62.524 -> 16006.261 MByte/s log_avg of all random : 0.041 0.650 7.255 34.331 55.544 41.600 || 24.679 -> 6317.766 MByte/s log_avg(ring,random) : 0.046 0.707 7.977 38.358 91.636 88.821 || 39.281 -> 10056.033 MByte/s SECTION-BY-PATTERN-MSGLNG-END SECTION-BEFF-BEGIN The effective bandwidth is b_eff = 10056.033 MByte/s on 256 processes ( = 39.281 MByte/s * 256 processes) system parameters : 256 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 = 10056.033 MB/s = 39.281 * 256 PEs with 128 MB/PE on sn6715 hwwt3e 2.0.4.71 unicosmk CRAY T3E