b_eff = 355.045 MB/s = 88.761 * 4 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 = 4 1-dim-paterns: size = 4 2-dim-paterns: size = 2 * 2 3-dim-paterns: size = 2 * 2 * 1 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 = 91.089 sec sum of max elapsed time per entries above = 91.917 sec difference = -0.828 sec = 0.9% The difference is less than 5 % SECTION-LOOP-END SECTION-PATTERN-BEGIN Pattern parameters: p00 ring-2*2fix => 1 sendrecv_calls with 4 messages, i.e. msgs/used node, all nodes are used p01 ring-1*4fix => 2 sendrecv_calls with 8 messages, i.e. msgs/used node, all nodes are used p02 ring-1*4fix => 2 sendrecv_calls with 8 messages, i.e. msgs/used node, all nodes are used p03 ring-1*4fix => 2 sendrecv_calls with 8 messages, i.e. msgs/used node, all nodes are used p04 ring-1*4fix => 2 sendrecv_calls with 8 messages, i.e. msgs/used node, all nodes are used p05 ring-1*4fix => 2 sendrecv_calls with 8 messages, i.e. msgs/used node, all nodes are used p06 random-cyc-1dim => 2 sendrecv_calls with 8 messages, i.e. msgs/used node, all nodes are used p07 random-cyc-1dim => 2 sendrecv_calls with 8 messages, i.e. msgs/used node, all nodes are used p08 random-cyc-1dim => 2 sendrecv_calls with 8 messages, i.e. msgs/used node, all nodes are used p09 random-cyc-1dim => 2 sendrecv_calls with 8 messages, i.e. msgs/used node, all nodes are used p10 random-cyc-1dim => 2 sendrecv_calls with 8 messages, i.e. msgs/used node, all nodes are used p11 random-cyc-1dim => 2 sendrecv_calls with 8 messages, i.e. msgs/used node, all nodes are used p12 random-cyc-1dim => 2 sendrecv_calls with 8 messages, i.e. msgs/used node, all nodes are used p13 random-cyc-1dim => 2 sendrecv_calls with 8 messages, i.e. msgs/used node, all nodes are used p14 random-cyc-1dim => 2 sendrecv_calls with 8 messages, i.e. msgs/used node, all nodes are used p15 random-cyc-1dim => 2 sendrecv_calls with 8 messages, i.e. msgs/used node, all nodes are used p16 worst-cyc-1dim => 2 sendrecv_calls with 8 messages, i.e. msgs/used node, all nodes are used p17 best bi-section => 2 sendrecv_calls with 4 messages, i.e. msgs/used node, all nodes are used p18 worst bi-section => 2 sendrecv_calls with 4 messages, i.e. msgs/used node, all nodes are used p19 acyclic-1dim-all => 2 sendrecv_calls with 6 messages, i.e. msgs/used node, all nodes are used p20 acyclic-2dim-all => 4 sendrecv_calls with 8 messages, i.e. msgs/used node, all nodes are used p21 acyclic-3dim-all => 4 sendrecv_calls with 8 messages, i.e. msgs/used node, all nodes are used p22 cyclic-1dim-all => 2 sendrecv_calls with 8 messages, i.e. msgs/used node, all nodes are used p23 cyclic-2dim-all => 2 sendrecv_calls with 8 messages, i.e. msgs/used node, all nodes are used p24 cyclic-3dim-all => 2 sendrecv_calls with 8 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-2*2fix : 72.599 90.536 65.746 -> 90.536 -> 362.143 MByte/s p01 ring-1*4fix : 65.861 88.835 52.165 -> 88.835 -> 355.341 MByte/s p02 ring-1*4fix : 65.829 88.824 53.009 -> 88.824 -> 355.297 MByte/s p03 ring-1*4fix : 66.028 88.733 50.821 -> 88.733 -> 354.932 MByte/s p04 ring-1*4fix : 64.788 88.602 50.961 -> 88.602 -> 354.407 MByte/s p05 ring-1*4fix : 65.917 88.695 50.832 -> 88.695 -> 354.779 MByte/s p06 random-cyc-1dim : 65.982 88.663 50.548 -> 88.663 -> 354.652 MByte/s p07 random-cyc-1dim : 66.825 87.971 51.279 -> 87.971 -> 351.882 MByte/s p08 random-cyc-1dim : 64.375 88.827 51.436 -> 88.827 -> 355.308 MByte/s p09 random-cyc-1dim : 64.711 88.423 52.433 -> 88.423 -> 353.692 MByte/s p10 random-cyc-1dim : 64.370 88.768 49.403 -> 88.768 -> 355.073 MByte/s p11 random-cyc-1dim : 67.370 88.050 52.060 -> 88.050 -> 352.201 MByte/s p12 random-cyc-1dim : 64.156 88.660 50.187 -> 88.660 -> 354.638 MByte/s p13 random-cyc-1dim : 67.540 87.865 51.574 -> 87.865 -> 351.459 MByte/s p14 random-cyc-1dim : 64.110 88.714 51.354 -> 88.714 -> 354.854 MByte/s p15 random-cyc-1dim : 64.180 88.625 50.229 -> 88.625 -> 354.502 MByte/s -- additional patterns that are not used to compute b_eff: p16 worst-cyc-1dim : 68.339 88.006 52.198 -> 88.006 -> 352.022 MByte/s p17 best bi-section : 51.925 90.266 50.619 -> 90.266 -> 361.064 MByte/s p18 worst bi-section : 42.218 68.688 44.218 -> 68.688 -> 274.753 MByte/s p19 acyclic-1dim-all : 53.348 62.512 41.178 -> 62.512 -> 250.047 MByte/s p20 acyclic-2dim-all : 51.981 58.381 50.465 -> 58.381 -> 233.525 MByte/s p21 acyclic-3dim-all : 51.952 58.946 54.066 -> 58.946 -> 235.783 MByte/s p22 cyclic-1dim-all : 64.417 87.785 51.225 -> 87.785 -> 351.141 MByte/s p23 cyclic-2dim-all : 68.857 57.480 54.732 -> 68.857 -> 275.429 MByte/s p24 cyclic-3dim-all : 69.541 60.421 57.703 -> 69.541 -> 278.165 MByte/s log_avg of all rings : 66.788 89.035 53.683 || 89.035 -> 356.140 MByte/s log_avg of all random : 65.348 88.456 51.042 || 88.456 -> 353.823 MByte/s log_avg(ring,random) : 66.064 88.745 52.346 ||( 88.745 -> 354.980)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-2*2fix : 90.246 90.190 90.231 -> 90.246 -> 360.982 MByte/s p01 ring-1*4fix : 88.686 88.562 88.702 -> 88.702 -> 354.808 MByte/s p02 ring-1*4fix : 88.615 88.380 88.522 -> 88.615 -> 354.461 MByte/s p03 ring-1*4fix : 88.323 88.616 88.157 -> 88.616 -> 354.463 MByte/s p04 ring-1*4fix : 88.161 88.502 88.158 -> 88.502 -> 354.009 MByte/s p05 ring-1*4fix : 88.424 88.393 88.229 -> 88.424 -> 353.697 MByte/s p06 random-cyc-1dim : 86.520 88.362 88.577 -> 88.577 -> 354.308 MByte/s p07 random-cyc-1dim : 87.396 87.368 87.811 -> 87.811 -> 351.246 MByte/s p08 random-cyc-1dim : 87.703 88.608 88.468 -> 88.608 -> 354.433 MByte/s p09 random-cyc-1dim : 88.278 88.256 88.210 -> 88.278 -> 353.112 MByte/s p10 random-cyc-1dim : 88.384 88.517 88.372 -> 88.517 -> 354.068 MByte/s p11 random-cyc-1dim : 87.554 87.414 87.082 -> 87.554 -> 350.214 MByte/s p12 random-cyc-1dim : 87.858 87.530 88.277 -> 88.277 -> 353.106 MByte/s p13 random-cyc-1dim : 87.609 87.778 87.172 -> 87.778 -> 351.113 MByte/s p14 random-cyc-1dim : 88.268 88.444 88.406 -> 88.444 -> 353.775 MByte/s p15 random-cyc-1dim : 88.364 88.287 88.466 -> 88.466 -> 353.863 MByte/s -- additional patterns that are not used to compute b_eff: p16 worst-cyc-1dim : 87.883 87.602 87.418 -> 87.883 -> 351.531 MByte/s p17 best bi-section : 89.794 89.686 90.009 -> 90.009 -> 360.035 MByte/s p18 worst bi-section : 68.442 68.407 68.597 -> 68.597 -> 274.387 MByte/s p19 acyclic-1dim-all : 65.014 65.013 65.688 -> 65.688 -> 262.753 MByte/s p20 acyclic-2dim-all : 66.044 65.635 66.306 -> 66.306 -> 265.225 MByte/s p21 acyclic-3dim-all : 70.254 66.118 66.441 -> 70.254 -> 281.017 MByte/s p22 cyclic-1dim-all : 87.273 87.639 87.610 -> 87.639 -> 350.557 MByte/s p23 cyclic-2dim-all : 71.061 71.894 72.532 -> 72.532 -> 290.130 MByte/s p24 cyclic-3dim-all : 71.531 72.195 74.587 -> 74.587 -> 298.348 MByte/s log_avg of all rings : 88.740 88.772 88.663 || 88.849 -> 355.394 MByte/s log_avg of all random : 87.792 88.055 88.083 || 88.230 -> 352.921 MByte/s log_avg(ring,random) : 88.264 88.413 88.373 ||( 88.539 -> 354.155)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-2*2fix p00 method 0 : 0.056 0.877 10.534 52.340 180.440 208.498 -> 72.599 -> 290.395 MByte/s p00 method 1 : 0.062 1.274 17.790 123.859 201.141 209.911 -> 90.536 -> 362.143 MByte/s p00 method 2 : 0.025 0.405 5.437 27.173 168.711 208.064 -> 65.746 -> 262.984 MByte/s p01 ring-1*4fix p01 method 0 : 0.055 0.859 10.154 50.930 173.063 155.293 -> 65.861 -> 263.443 MByte/s p01 method 1 : 0.074 1.437 19.263 125.854 194.882 200.485 -> 88.835 -> 355.341 MByte/s p01 method 2 : 0.025 0.398 4.764 29.873 132.695 154.046 -> 52.165 -> 208.662 MByte/s p02 ring-1*4fix p02 method 0 : 0.055 0.867 10.137 50.389 173.168 155.254 -> 65.829 -> 263.316 MByte/s p02 method 1 : 0.073 1.443 19.502 126.200 195.345 199.948 -> 88.824 -> 355.297 MByte/s p02 method 2 : 0.023 0.368 5.103 29.772 133.919 152.821 -> 53.009 -> 212.035 MByte/s p03 ring-1*4fix p03 method 0 : 0.056 0.868 10.238 50.910 173.186 155.452 -> 66.028 -> 264.111 MByte/s p03 method 1 : 0.074 1.431 19.213 125.925 195.094 200.141 -> 88.733 -> 354.932 MByte/s p03 method 2 : 0.023 0.378 5.130 29.991 129.275 126.687 -> 50.821 -> 203.284 MByte/s p04 ring-1*4fix p04 method 0 : 0.056 0.879 10.294 51.068 173.209 155.434 -> 64.788 -> 259.151 MByte/s p04 method 1 : 0.073 1.436 19.178 125.169 194.968 200.171 -> 88.602 -> 354.407 MByte/s p04 method 2 : 0.022 0.371 4.756 29.955 127.931 119.781 -> 50.961 -> 203.845 MByte/s p05 ring-1*4fix p05 method 0 : 0.056 0.875 9.979 51.102 173.129 155.179 -> 65.917 -> 263.670 MByte/s p05 method 1 : 0.074 1.433 19.142 125.342 195.519 199.546 -> 88.695 -> 354.779 MByte/s p05 method 2 : 0.024 0.384 5.101 29.923 135.636 115.877 -> 50.832 -> 203.329 MByte/s p06 random-cyc-1dim p06 method 0 : 0.056 0.877 10.311 50.982 173.233 155.330 -> 65.982 -> 263.929 MByte/s p06 method 1 : 0.074 1.424 19.244 125.819 195.133 199.196 -> 88.663 -> 354.652 MByte/s p06 method 2 : 0.022 0.386 4.969 30.261 121.951 154.301 -> 50.548 -> 202.192 MByte/s p07 random-cyc-1dim p07 method 0 : 0.056 0.873 10.306 51.078 173.344 141.187 -> 66.825 -> 267.302 MByte/s p07 method 1 : 0.074 1.419 19.155 126.259 195.856 195.471 -> 87.971 -> 351.882 MByte/s p07 method 2 : 0.025 0.372 4.848 30.258 134.606 145.771 -> 51.279 -> 205.114 MByte/s p08 random-cyc-1dim p08 method 0 : 0.056 0.877 10.309 50.931 173.020 155.479 -> 64.375 -> 257.500 MByte/s p08 method 1 : 0.074 1.432 19.353 126.518 195.519 200.842 -> 88.827 -> 355.308 MByte/s p08 method 2 : 0.024 0.387 4.696 30.097 150.501 126.421 -> 51.436 -> 205.744 MByte/s p09 random-cyc-1dim p09 method 0 : 0.056 0.872 10.349 50.971 173.157 154.304 -> 64.711 -> 258.843 MByte/s p09 method 1 : 0.078 1.438 19.120 125.595 194.847 196.238 -> 88.423 -> 353.692 MByte/s p09 method 2 : 0.023 0.397 4.742 30.195 139.086 121.876 -> 52.433 -> 209.730 MByte/s p10 random-cyc-1dim p10 method 0 : 0.056 0.875 10.251 51.068 173.645 126.230 -> 64.370 -> 257.481 MByte/s p10 method 1 : 0.077 1.426 19.221 124.974 195.419 202.885 -> 88.768 -> 355.073 MByte/s p10 method 2 : 0.024 0.404 5.101 30.024 127.445 129.250 -> 49.403 -> 197.610 MByte/s p11 random-cyc-1dim p11 method 0 : 0.056 0.877 10.352 50.551 173.656 141.923 -> 67.370 -> 269.482 MByte/s p11 method 1 : 0.076 1.438 19.105 126.096 196.153 201.364 -> 88.050 -> 352.201 MByte/s p11 method 2 : 0.025 0.376 4.949 29.855 124.786 138.364 -> 52.060 -> 208.240 MByte/s p12 random-cyc-1dim p12 method 0 : 0.056 0.876 10.261 50.921 172.828 155.952 -> 64.156 -> 256.622 MByte/s p12 method 1 : 0.077 1.430 19.141 125.777 195.089 199.406 -> 88.660 -> 354.638 MByte/s p12 method 2 : 0.023 0.398 4.841 30.254 129.114 139.659 -> 50.187 -> 200.747 MByte/s p13 random-cyc-1dim p13 method 0 : 0.056 0.879 10.377 50.863 173.510 145.235 -> 67.540 -> 270.160 MByte/s p13 method 1 : 0.077 1.431 19.243 126.342 194.908 200.588 -> 87.865 -> 351.459 MByte/s p13 method 2 : 0.024 0.376 4.697 30.162 128.366 121.271 -> 51.574 -> 206.295 MByte/s p14 random-cyc-1dim p14 method 0 : 0.056 0.877 10.337 50.945 173.278 131.070 -> 64.110 -> 256.438 MByte/s p14 method 1 : 0.077 1.443 19.261 124.925 195.204 201.701 -> 88.714 -> 354.854 MByte/s p14 method 2 : 0.023 0.374 4.772 29.883 126.700 152.497 -> 51.354 -> 205.416 MByte/s p15 random-cyc-1dim p15 method 0 : 0.055 0.881 10.344 50.849 173.034 155.495 -> 64.180 -> 256.721 MByte/s p15 method 1 : 0.077 1.423 19.430 125.809 195.167 197.089 -> 88.625 -> 354.502 MByte/s p15 method 2 : 0.024 0.373 5.074 29.690 125.472 113.807 -> 50.229 -> 200.915 MByte/s -- additional patterns that are not used to compute b_eff: p16 worst-cyc-1dim p16 method 0 : 0.056 0.881 10.353 51.096 173.786 160.840 -> 68.339 -> 273.358 MByte/s p16 method 1 : 0.077 1.431 19.258 126.089 195.044 202.081 -> 88.006 -> 352.022 MByte/s p16 method 2 : 0.024 0.398 4.701 29.963 127.255 143.621 -> 52.198 -> 208.790 MByte/s p17 best bi-section p17 method 0 : 0.031 0.481 5.594 28.397 131.674 163.057 -> 51.925 -> 207.701 MByte/s p17 method 1 : 0.065 1.242 17.120 121.513 201.228 210.836 -> 90.266 -> 361.064 MByte/s p17 method 2 : 0.015 0.250 3.508 28.226 129.005 162.224 -> 50.619 -> 202.478 MByte/s p18 worst bi-section p18 method 0 : 0.030 0.462 5.447 28.122 105.028 123.258 -> 42.218 -> 168.870 MByte/s p18 method 1 : 0.062 1.212 16.608 100.166 147.791 152.416 -> 68.688 -> 274.753 MByte/s p18 method 2 : 0.016 0.248 3.630 27.829 106.113 122.859 -> 44.218 -> 176.874 MByte/s p19 acyclic-1dim-all p19 method 0 : 0.039 0.616 7.434 38.196 132.267 165.521 -> 53.348 -> 213.393 MByte/s p19 method 1 : 0.059 1.119 14.936 92.833 136.351 125.546 -> 62.512 -> 250.047 MByte/s p19 method 2 : 0.017 0.260 3.583 25.285 103.708 109.354 -> 41.178 -> 164.711 MByte/s p20 acyclic-2dim-all p20 method 0 : 0.031 0.492 5.743 28.623 131.768 163.111 -> 51.981 -> 207.925 MByte/s p20 method 1 : 0.077 1.484 19.017 89.714 113.533 116.128 -> 58.381 -> 233.525 MByte/s p20 method 2 : 0.015 0.238 3.288 26.182 127.470 177.500 -> 50.465 -> 201.862 MByte/s p21 acyclic-3dim-all p21 method 0 : 0.031 0.481 5.740 28.391 132.002 162.983 -> 51.952 -> 207.809 MByte/s p21 method 1 : 0.076 1.477 18.704 111.845 113.664 116.184 -> 58.946 -> 235.783 MByte/s p21 method 2 : 0.015 0.238 3.294 26.640 137.597 183.248 -> 54.066 -> 216.265 MByte/s p22 cyclic-1dim-all p22 method 0 : 0.055 0.865 10.372 51.100 173.172 155.927 -> 64.417 -> 257.667 MByte/s p22 method 1 : 0.071 1.331 17.904 122.423 194.647 200.223 -> 87.785 -> 351.141 MByte/s p22 method 2 : 0.024 0.380 4.804 29.894 139.636 117.001 -> 51.225 -> 204.901 MByte/s p23 cyclic-2dim-all p23 method 0 : 0.057 0.886 10.418 51.887 178.114 183.216 -> 68.857 -> 275.429 MByte/s p23 method 1 : 0.073 1.375 18.561 111.429 113.499 116.070 -> 57.480 -> 229.920 MByte/s p23 method 2 : 0.026 0.364 4.672 31.019 150.484 160.172 -> 54.732 -> 218.929 MByte/s p24 cyclic-3dim-all p24 method 0 : 0.057 0.889 10.428 51.663 179.428 208.638 -> 69.541 -> 278.165 MByte/s p24 method 1 : 0.073 1.356 18.211 89.701 184.815 118.550 -> 60.421 -> 241.683 MByte/s p24 method 2 : 0.023 0.358 4.672 30.897 140.825 207.899 -> 57.703 -> 230.812 MByte/s log_avg of all rings - ring, method 0 : 0.056 0.871 10.221 51.120 174.345 163.134 || 66.788 -> 267.152 MByte/s - ring, method 1 : 0.072 1.408 19.006 125.389 196.145 201.667 || 89.035 -> 356.140 MByte/s - ring, method 2 : 0.024 0.383 5.043 29.430 137.389 143.172 || 53.683 -> 214.732 MByte/s log_avg of all random - random, method 0 : 0.056 0.876 10.320 50.916 173.270 145.834 || 65.348 -> 261.393 MByte/s - random, method 1 : 0.076 1.430 19.227 125.810 195.329 199.464 || 88.456 -> 353.823 MByte/s - random, method 2 : 0.024 0.384 4.867 30.067 130.566 133.675 || 51.042 -> 204.169 MByte/s log_avg(ring,random) - average, method 0 : 0.056 0.873 10.270 51.018 173.807 154.242 || 66.064 -> 264.257 MByte/s - average, method 1 : 0.074 1.419 19.116 125.600 195.737 200.563 || 88.745 -> 354.980 MByte/s - average, method 2 : 0.024 0.384 4.954 29.747 133.934 138.342 || 52.346 -> 209.384 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 0.295 0.074 0.072 0.076 0.056 0.074 0.024 2 0.593 0.148 0.144 0.153 0.112 0.148 0.049 4 1.512 0.378 0.373 0.383 0.224 0.378 0.099 8 3.001 0.750 0.740 0.760 0.446 0.750 0.198 16 5.676 1.419 1.408 1.430 0.873 1.419 0.384 32 10.854 2.714 2.681 2.746 1.428 2.714 0.695 64 19.303 4.826 4.749 4.903 2.594 4.826 1.267 128 43.294 10.823 10.739 10.908 5.629 10.823 2.623 256 76.465 19.116 19.006 19.227 10.270 19.116 4.954 512 135.969 33.992 33.842 34.144 17.886 33.992 8.791 1024 232.447 58.112 57.966 58.258 28.831 58.112 15.226 2048 351.560 87.890 87.772 88.008 38.961 87.890 23.347 4096 502.398 125.600 125.389 125.810 51.018 125.600 29.747 8192 625.234 156.308 156.563 156.055 87.399 156.308 77.684 16384 706.320 176.580 177.010 176.151 121.618 176.580 107.390 32768 755.311 188.828 189.386 188.272 152.032 188.828 125.970 65536 782.948 195.737 196.145 195.329 173.807 195.737 133.934 131072 791.673 197.918 199.615 196.236 186.426 197.918 142.375 262144 801.785 200.446 201.673 199.227 183.726 200.430 141.588 524288 806.637 201.659 202.661 200.662 167.917 201.327 141.249 1048576 802.251 200.563 201.667 199.464 154.242 200.563 138.342 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-2*2fix : 0.062 1.274 17.790 123.859 201.141 209.911 -> 90.536 -> 362.143 MByte/s p01 ring-1*4fix : 0.074 1.437 19.263 125.854 194.882 200.485 -> 88.835 -> 355.341 MByte/s p02 ring-1*4fix : 0.073 1.443 19.502 126.200 195.345 199.948 -> 88.824 -> 355.297 MByte/s p03 ring-1*4fix : 0.074 1.431 19.213 125.925 195.094 200.141 -> 88.733 -> 354.932 MByte/s p04 ring-1*4fix : 0.073 1.436 19.178 125.169 194.968 200.171 -> 88.602 -> 354.407 MByte/s p05 ring-1*4fix : 0.074 1.433 19.142 125.342 195.519 199.546 -> 88.695 -> 354.779 MByte/s p06 random-cyc-1dim : 0.074 1.424 19.244 125.819 195.133 199.196 -> 88.663 -> 354.652 MByte/s p07 random-cyc-1dim : 0.074 1.419 19.155 126.259 195.856 195.471 -> 87.997 -> 351.989 MByte/s p08 random-cyc-1dim : 0.074 1.432 19.353 126.518 195.519 200.842 -> 88.827 -> 355.308 MByte/s p09 random-cyc-1dim : 0.078 1.438 19.120 125.595 194.847 196.238 -> 88.423 -> 353.692 MByte/s p10 random-cyc-1dim : 0.077 1.426 19.221 124.974 195.419 202.885 -> 88.768 -> 355.073 MByte/s p11 random-cyc-1dim : 0.076 1.438 19.105 126.096 196.153 201.364 -> 88.050 -> 352.201 MByte/s p12 random-cyc-1dim : 0.077 1.430 19.141 125.777 195.089 199.406 -> 88.660 -> 354.638 MByte/s p13 random-cyc-1dim : 0.077 1.431 19.243 126.342 194.908 200.588 -> 88.162 -> 352.646 MByte/s p14 random-cyc-1dim : 0.077 1.443 19.261 124.925 195.204 201.701 -> 88.714 -> 354.854 MByte/s p15 random-cyc-1dim : 0.077 1.423 19.430 125.809 195.167 197.089 -> 88.625 -> 354.502 MByte/s -- additional patterns that are not used to compute b_eff: p16 worst-cyc-1dim : 0.077 1.431 19.258 126.089 195.044 202.081 -> 88.143 -> 352.571 MByte/s p17 best bi-section : 0.065 1.242 17.120 121.513 201.228 210.836 -> 90.266 -> 361.064 MByte/s p18 worst bi-section : 0.062 1.212 16.608 100.166 147.791 152.416 -> 68.688 -> 274.753 MByte/s p19 acyclic-1dim-all : 0.059 1.119 14.936 92.833 136.351 165.521 -> 66.310 -> 265.239 MByte/s p20 acyclic-2dim-all : 0.077 1.484 19.017 89.714 131.768 177.500 -> 67.806 -> 271.224 MByte/s p21 acyclic-3dim-all : 0.076 1.477 18.704 111.845 137.597 183.248 -> 70.562 -> 282.250 MByte/s p22 cyclic-1dim-all : 0.071 1.331 17.904 122.423 194.647 200.223 -> 87.785 -> 351.141 MByte/s p23 cyclic-2dim-all : 0.073 1.375 18.561 111.429 178.114 183.216 -> 76.455 -> 305.821 MByte/s p24 cyclic-3dim-all : 0.073 1.356 18.211 89.701 184.815 208.638 -> 76.831 -> 307.326 MByte/s log_avg of all rings : 0.072 1.408 19.006 125.389 196.145 201.667 || 89.035 -> 356.140 MByte/s log_avg of all random : 0.076 1.430 19.227 125.810 195.329 199.464 || 88.488 -> 353.954 MByte/s log_avg(ring,random) : 0.074 1.419 19.116 125.600 195.737 200.563 || 88.761 -> 355.045 MByte/s SECTION-BY-PATTERN-MSGLNG-END SECTION-BEFF-BEGIN The effective bandwidth is b_eff = 355.045 MByte/s on 4 processes ( = 88.761 MByte/s * 4 processes) system parameters : 4 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 = 355.045 MB/s = 88.761 * 4 PEs with 128 MB/PE on sn6715 hwwt3e 2.0.4.71 unicosmk CRAY T3E