b_eff = 276.903 MB/s = 34.613 * 8 PEs with 256 MB/PE on HI-UX/MPP hitachi 02-03 0 SR2201 SECTION-HEAD-BEGIN b_eff.c, Revision 3.2 from Nov. 07, 1999 MEMORY_PER_PROCESSOR = 256 MBytes [1M = 1024*1024] 1-dim-paterns: size = 8 1-dim-paterns: size = 8 2-dim-paterns: size = 4 * 2 3-dim-paterns: size = 2 * 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 (300), 8933 (234), 19484 (107), 42495 (49), 92682 (22), 202141 (10), 440872 (4), 961548 (2), 2097152 (1), SECTION-HEAD-END SECTION-LOOP-BEGIN measurment loop: elapsed time = 399.805 sec sum of max elapsed time per entries above = 400.261 sec difference = -0.456 sec = 0.1% The difference is less than 5 % SECTION-LOOP-END SECTION-PATTERN-BEGIN Pattern parameters: p00 ring-4*2fix => 1 sendrecv_calls with 8 messages, i.e. msgs/used node, all nodes are used p01 ring-2*4fix => 2 sendrecv_calls with 16 messages, i.e. msgs/used node, all nodes are used p02 ring-1*8fix => 2 sendrecv_calls with 16 messages, i.e. msgs/used node, all nodes are used p03 ring-1*8fix => 2 sendrecv_calls with 16 messages, i.e. msgs/used node, all nodes are used p04 ring-1*8fix => 2 sendrecv_calls with 16 messages, i.e. msgs/used node, all nodes are used p05 ring-1*8fix => 2 sendrecv_calls with 16 messages, i.e. msgs/used node, all nodes are used p06 random-cyc-1dim => 2 sendrecv_calls with 16 messages, i.e. msgs/used node, all nodes are used p07 random-cyc-1dim => 2 sendrecv_calls with 16 messages, i.e. msgs/used node, all nodes are used p08 random-cyc-1dim => 2 sendrecv_calls with 16 messages, i.e. msgs/used node, all nodes are used p09 random-cyc-1dim => 2 sendrecv_calls with 16 messages, i.e. msgs/used node, all nodes are used p10 random-cyc-1dim => 2 sendrecv_calls with 16 messages, i.e. msgs/used node, all nodes are used p11 random-cyc-1dim => 2 sendrecv_calls with 16 messages, i.e. msgs/used node, all nodes are used p12 random-cyc-1dim => 2 sendrecv_calls with 16 messages, i.e. msgs/used node, all nodes are used p13 random-cyc-1dim => 2 sendrecv_calls with 16 messages, i.e. msgs/used node, all nodes are used p14 random-cyc-1dim => 2 sendrecv_calls with 16 messages, i.e. msgs/used node, all nodes are used p15 random-cyc-1dim => 2 sendrecv_calls with 16 messages, i.e. msgs/used node, all nodes are used p16 worst-cyc-1dim => 2 sendrecv_calls with 16 messages, i.e. msgs/used node, all nodes are used p17 best bi-section => 2 sendrecv_calls with 8 messages, i.e. msgs/used node, all nodes are used p18 worst bi-section => 2 sendrecv_calls with 8 messages, i.e. msgs/used node, all nodes are used p19 acyclic-1dim-all => 2 sendrecv_calls with 14 messages, i.e. msgs/used node, all nodes are used p20 acyclic-2dim-all => 4 sendrecv_calls with 20 messages, i.e. msgs/used node, all nodes are used p21 acyclic-3dim-all => 6 sendrecv_calls with 24 messages, i.e. msgs/used node, all nodes are used p22 cyclic-1dim-all => 2 sendrecv_calls with 16 messages, i.e. msgs/used node, all nodes are used p23 cyclic-2dim-all => 3 sendrecv_calls with 24 messages, i.e. msgs/used node, all nodes are used p24 cyclic-3dim-all => 3 sendrecv_calls with 24 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-4*2fix : 39.144 16.253 36.038 -> 39.144 -> 313.149 MByte/s p01 ring-2*4fix : 33.473 23.088 33.368 -> 33.473 -> 267.788 MByte/s p02 ring-1*8fix : 33.191 18.120 33.192 -> 33.192 -> 265.537 MByte/s p03 ring-1*8fix : 32.991 18.118 33.488 -> 33.488 -> 267.905 MByte/s p04 ring-1*8fix : 33.192 18.091 33.538 -> 33.538 -> 268.300 MByte/s p05 ring-1*8fix : 32.987 18.103 33.701 -> 33.701 -> 269.604 MByte/s p06 random-cyc-1dim : 32.923 18.080 33.259 -> 33.259 -> 266.071 MByte/s p07 random-cyc-1dim : 32.793 18.044 33.451 -> 33.451 -> 267.605 MByte/s p08 random-cyc-1dim : 32.856 18.107 33.269 -> 33.269 -> 266.154 MByte/s p09 random-cyc-1dim : 33.134 18.100 33.428 -> 33.428 -> 267.422 MByte/s p10 random-cyc-1dim : 33.150 18.115 32.961 -> 33.150 -> 265.198 MByte/s p11 random-cyc-1dim : 32.894 18.130 32.874 -> 32.894 -> 263.152 MByte/s p12 random-cyc-1dim : 33.464 18.104 33.674 -> 33.674 -> 269.395 MByte/s p13 random-cyc-1dim : 33.103 18.084 33.160 -> 33.160 -> 265.281 MByte/s p14 random-cyc-1dim : 33.104 18.113 33.237 -> 33.237 -> 265.897 MByte/s p15 random-cyc-1dim : 32.848 18.013 33.036 -> 33.036 -> 264.288 MByte/s -- additional patterns that are not used to compute b_eff: p16 worst-cyc-1dim : 33.462 18.025 33.111 -> 33.462 -> 267.695 MByte/s p17 best bi-section : 32.162 16.420 37.336 -> 37.336 -> 298.684 MByte/s p18 worst bi-section : 31.928 27.492 37.234 -> 37.234 -> 297.871 MByte/s p19 acyclic-1dim-all : 29.036 17.512 29.818 -> 29.818 -> 238.544 MByte/s p20 acyclic-2dim-all : 27.476 19.019 29.228 -> 29.228 -> 233.827 MByte/s p21 acyclic-3dim-all : 32.606 27.089 32.019 -> 32.606 -> 260.848 MByte/s p22 cyclic-1dim-all : 33.477 18.145 33.413 -> 33.477 -> 267.817 MByte/s p23 cyclic-2dim-all : 34.959 19.460 32.987 -> 34.959 -> 279.669 MByte/s p24 cyclic-3dim-all : 39.007 27.013 32.832 -> 39.007 -> 312.057 MByte/s log_avg of all rings : 34.095 18.520 33.874 || 34.362 -> 274.894 MByte/s log_avg of all random : 33.026 18.089 33.234 || 33.255 -> 266.041 MByte/s log_avg(ring,random) : 33.556 18.303 33.552 ||( 33.804 -> 270.431)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-4*2fix : 37.917 39.139 39.223 -> 39.223 -> 313.781 MByte/s p01 ring-2*4fix : 34.303 34.794 34.617 -> 34.794 -> 278.354 MByte/s p02 ring-1*8fix : 33.761 33.463 33.635 -> 33.761 -> 270.092 MByte/s p03 ring-1*8fix : 33.710 34.027 33.509 -> 34.027 -> 272.219 MByte/s p04 ring-1*8fix : 33.363 33.640 33.742 -> 33.742 -> 269.936 MByte/s p05 ring-1*8fix : 33.665 33.569 33.558 -> 33.665 -> 269.319 MByte/s p06 random-cyc-1dim : 33.133 33.345 33.827 -> 33.827 -> 270.615 MByte/s p07 random-cyc-1dim : 32.844 33.617 33.544 -> 33.617 -> 268.936 MByte/s p08 random-cyc-1dim : 33.202 33.597 33.560 -> 33.597 -> 268.775 MByte/s p09 random-cyc-1dim : 33.835 33.845 33.338 -> 33.845 -> 270.759 MByte/s p10 random-cyc-1dim : 34.013 33.281 33.453 -> 34.013 -> 272.102 MByte/s p11 random-cyc-1dim : 33.785 33.544 33.608 -> 33.785 -> 270.283 MByte/s p12 random-cyc-1dim : 33.752 33.789 34.153 -> 34.153 -> 273.226 MByte/s p13 random-cyc-1dim : 33.504 33.788 33.608 -> 33.788 -> 270.304 MByte/s p14 random-cyc-1dim : 33.856 33.571 33.862 -> 33.862 -> 270.897 MByte/s p15 random-cyc-1dim : 33.456 33.433 33.765 -> 33.765 -> 270.118 MByte/s -- additional patterns that are not used to compute b_eff: p16 worst-cyc-1dim : 33.231 34.176 33.831 -> 34.176 -> 273.404 MByte/s p17 best bi-section : 36.637 37.311 36.652 -> 37.311 -> 298.485 MByte/s p18 worst bi-section : 36.781 36.889 36.732 -> 36.889 -> 295.113 MByte/s p19 acyclic-1dim-all : 30.002 29.779 29.881 -> 30.002 -> 240.017 MByte/s p20 acyclic-2dim-all : 30.139 29.930 29.924 -> 30.139 -> 241.111 MByte/s p21 acyclic-3dim-all : 34.181 34.440 34.480 -> 34.480 -> 275.838 MByte/s p22 cyclic-1dim-all : 33.882 34.118 34.082 -> 34.118 -> 272.948 MByte/s p23 cyclic-2dim-all : 35.498 35.560 35.437 -> 35.560 -> 284.478 MByte/s p24 cyclic-3dim-all : 39.067 39.366 39.323 -> 39.366 -> 314.932 MByte/s log_avg of all rings : 34.419 34.718 34.657 || 34.816 -> 278.525 MByte/s log_avg of all random : 33.536 33.580 33.671 || 33.825 -> 270.599 MByte/s log_avg(ring,random) : 33.975 34.144 34.161 ||( 34.317 -> 274.533)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 92682 2097152 -> average -> accumulated p00 ring-4*2fix p00 method 0 : 0.016 0.255 2.978 33.708 103.713 107.773 -> 39.144 -> 313.149 MByte/s p00 method 1 : 0.002 0.033 0.510 7.279 44.094 53.820 -> 16.253 -> 130.023 MByte/s p00 method 2 : 0.018 0.270 3.152 34.967 81.665 100.961 -> 36.038 -> 288.307 MByte/s p01 ring-2*4fix p01 method 0 : 0.016 0.255 3.002 33.830 79.060 102.225 -> 33.473 -> 267.788 MByte/s p01 method 1 : 0.004 0.066 0.967 13.097 62.829 71.898 -> 23.088 -> 184.708 MByte/s p01 method 2 : 0.019 0.288 3.240 35.875 86.179 88.312 -> 33.368 -> 266.940 MByte/s p02 ring-1*8fix p02 method 0 : 0.017 0.255 2.955 33.797 81.657 100.764 -> 33.191 -> 265.526 MByte/s p02 method 1 : 0.004 0.066 0.930 12.109 49.178 53.781 -> 18.120 -> 144.958 MByte/s p02 method 2 : 0.019 0.288 3.260 35.822 84.588 90.119 -> 33.192 -> 265.537 MByte/s p03 ring-1*8fix p03 method 0 : 0.016 0.256 2.980 33.751 82.665 94.707 -> 32.991 -> 263.925 MByte/s p03 method 1 : 0.004 0.066 0.932 12.033 49.153 53.830 -> 18.118 -> 144.945 MByte/s p03 method 2 : 0.019 0.287 3.256 35.866 85.678 91.759 -> 33.488 -> 267.905 MByte/s p04 ring-1*8fix p04 method 0 : 0.016 0.256 2.939 33.860 83.998 93.610 -> 33.192 -> 265.538 MByte/s p04 method 1 : 0.004 0.067 0.928 12.072 49.084 53.411 -> 18.091 -> 144.725 MByte/s p04 method 2 : 0.019 0.287 3.250 35.842 85.820 88.686 -> 33.538 -> 268.300 MByte/s p05 ring-1*8fix p05 method 0 : 0.017 0.253 2.975 33.936 82.477 92.922 -> 32.987 -> 263.895 MByte/s p05 method 1 : 0.004 0.066 0.930 12.043 49.149 53.798 -> 18.103 -> 144.827 MByte/s p05 method 2 : 0.019 0.287 3.253 35.771 85.744 94.138 -> 33.701 -> 269.604 MByte/s p06 random-cyc-1dim p06 method 0 : 0.016 0.254 2.976 33.904 82.135 101.744 -> 32.923 -> 263.386 MByte/s p06 method 1 : 0.004 0.067 0.933 12.103 48.926 53.763 -> 18.080 -> 144.642 MByte/s p06 method 2 : 0.019 0.286 3.277 35.672 84.971 91.349 -> 33.259 -> 266.071 MByte/s p07 random-cyc-1dim p07 method 0 : 0.017 0.255 2.975 33.874 82.084 94.517 -> 32.793 -> 262.340 MByte/s p07 method 1 : 0.004 0.066 0.929 12.069 48.878 53.733 -> 18.044 -> 144.349 MByte/s p07 method 2 : 0.019 0.288 3.279 35.865 85.428 87.933 -> 33.451 -> 267.605 MByte/s p08 random-cyc-1dim p08 method 0 : 0.017 0.255 2.930 33.985 82.686 93.537 -> 32.856 -> 262.847 MByte/s p08 method 1 : 0.004 0.066 0.931 12.129 49.111 54.198 -> 18.107 -> 144.855 MByte/s p08 method 2 : 0.019 0.289 3.249 35.930 84.868 87.039 -> 33.269 -> 266.154 MByte/s p09 random-cyc-1dim p09 method 0 : 0.017 0.255 3.010 33.794 80.358 99.499 -> 33.134 -> 265.075 MByte/s p09 method 1 : 0.004 0.066 0.935 12.095 49.131 54.053 -> 18.100 -> 144.798 MByte/s p09 method 2 : 0.019 0.287 3.272 35.584 85.118 89.912 -> 33.428 -> 267.422 MByte/s p10 random-cyc-1dim p10 method 0 : 0.017 0.256 2.972 33.820 83.460 96.828 -> 33.150 -> 265.198 MByte/s p10 method 1 : 0.004 0.066 0.936 12.104 49.179 53.584 -> 18.115 -> 144.922 MByte/s p10 method 2 : 0.019 0.287 3.273 35.799 85.932 86.682 -> 32.961 -> 263.688 MByte/s p11 random-cyc-1dim p11 method 0 : 0.017 0.254 2.957 33.989 82.001 94.954 -> 32.894 -> 263.152 MByte/s p11 method 1 : 0.004 0.066 0.932 12.105 49.100 54.191 -> 18.130 -> 145.043 MByte/s p11 method 2 : 0.019 0.287 3.249 35.859 86.748 85.766 -> 32.874 -> 262.989 MByte/s p12 random-cyc-1dim p12 method 0 : 0.017 0.257 2.960 33.709 84.204 101.486 -> 33.464 -> 267.713 MByte/s p12 method 1 : 0.004 0.067 0.934 12.105 48.963 53.992 -> 18.104 -> 144.835 MByte/s p12 method 2 : 0.019 0.287 3.256 35.780 87.744 93.230 -> 33.674 -> 269.395 MByte/s p13 random-cyc-1dim p13 method 0 : 0.017 0.255 2.968 33.932 82.812 97.749 -> 33.103 -> 264.825 MByte/s p13 method 1 : 0.004 0.066 0.931 12.079 49.079 53.741 -> 18.084 -> 144.671 MByte/s p13 method 2 : 0.019 0.288 3.270 35.746 84.326 90.036 -> 33.160 -> 265.281 MByte/s p14 random-cyc-1dim p14 method 0 : 0.016 0.253 2.951 33.728 81.586 93.870 -> 33.104 -> 264.832 MByte/s p14 method 1 : 0.004 0.066 0.924 12.043 49.083 54.067 -> 18.113 -> 144.907 MByte/s p14 method 2 : 0.019 0.278 3.205 35.231 84.923 88.413 -> 33.237 -> 265.897 MByte/s p15 random-cyc-1dim p15 method 0 : 0.016 0.250 2.955 33.604 81.881 96.215 -> 32.848 -> 262.786 MByte/s p15 method 1 : 0.004 0.064 0.914 11.914 48.992 53.828 -> 18.013 -> 144.107 MByte/s p15 method 2 : 0.019 0.285 3.246 35.466 87.174 86.189 -> 33.036 -> 264.288 MByte/s -- additional patterns that are not used to compute b_eff: p16 worst-cyc-1dim p16 method 0 : 0.016 0.250 2.918 33.528 81.622 101.410 -> 33.462 -> 267.695 MByte/s p16 method 1 : 0.004 0.066 0.921 11.977 48.939 53.657 -> 18.025 -> 144.197 MByte/s p16 method 2 : 0.019 0.281 3.255 35.509 85.209 84.908 -> 33.111 -> 264.887 MByte/s p17 best bi-section p17 method 0 : 0.018 0.280 2.520 27.645 78.315 104.926 -> 32.162 -> 257.294 MByte/s p17 method 1 : 0.002 0.033 0.504 7.181 44.107 56.430 -> 16.420 -> 131.360 MByte/s p17 method 2 : 0.014 0.212 2.590 30.250 100.870 111.521 -> 37.336 -> 298.684 MByte/s p18 worst bi-section p18 method 0 : 0.018 0.273 2.512 27.285 78.387 102.600 -> 31.928 -> 255.423 MByte/s p18 method 1 : 0.002 0.033 0.505 7.502 70.420 106.303 -> 27.492 -> 219.936 MByte/s p18 method 2 : 0.014 0.212 2.598 30.219 98.445 112.129 -> 37.234 -> 297.871 MByte/s p19 acyclic-1dim-all p19 method 0 : 0.014 0.223 2.614 29.642 72.814 80.284 -> 29.036 -> 232.291 MByte/s p19 method 1 : 0.004 0.057 0.828 10.967 48.508 48.112 -> 17.512 -> 140.099 MByte/s p19 method 2 : 0.017 0.253 2.856 31.802 75.048 77.270 -> 29.818 -> 238.544 MByte/s p20 acyclic-2dim-all p20 method 0 : 0.014 0.216 2.278 25.882 66.615 86.757 -> 27.476 -> 219.807 MByte/s p20 method 1 : 0.005 0.082 1.141 14.283 50.908 55.167 -> 19.019 -> 152.156 MByte/s p20 method 2 : 0.015 0.229 2.572 28.718 77.652 70.161 -> 29.228 -> 233.827 MByte/s p21 acyclic-3dim-all p21 method 0 : 0.018 0.288 2.597 28.188 79.564 105.342 -> 32.606 -> 260.848 MByte/s p21 method 1 : 0.006 0.099 1.352 17.598 73.823 81.882 -> 27.089 -> 216.715 MByte/s p21 method 2 : 0.015 0.231 2.722 31.048 84.403 86.517 -> 32.019 -> 256.154 MByte/s p22 cyclic-1dim-all p22 method 0 : 0.016 0.262 3.016 33.615 83.525 95.708 -> 33.477 -> 267.817 MByte/s p22 method 1 : 0.004 0.067 0.932 12.001 49.165 54.028 -> 18.145 -> 145.161 MByte/s p22 method 2 : 0.018 0.298 3.280 35.219 85.387 88.902 -> 33.413 -> 267.307 MByte/s p23 cyclic-2dim-all p23 method 0 : 0.016 0.257 2.996 33.543 85.753 103.664 -> 34.959 -> 279.669 MByte/s p23 method 1 : 0.006 0.099 1.346 16.515 51.393 54.827 -> 19.460 -> 155.682 MByte/s p23 method 2 : 0.019 0.300 3.344 35.627 84.159 79.320 -> 32.987 -> 263.899 MByte/s p24 cyclic-3dim-all p24 method 0 : 0.017 0.257 2.984 33.831 103.183 106.706 -> 39.007 -> 312.057 MByte/s p24 method 1 : 0.006 0.099 1.377 18.039 73.354 80.209 -> 27.013 -> 216.102 MByte/s p24 method 2 : 0.019 0.303 3.364 35.787 83.525 87.905 -> 32.832 -> 262.654 MByte/s log_avg of all rings - ring, method 0 : 0.016 0.255 2.972 33.814 85.235 98.522 || 34.095 -> 272.760 MByte/s - ring, method 1 : 0.004 0.059 0.847 11.243 50.279 56.401 || 18.520 -> 148.158 MByte/s - ring, method 2 : 0.019 0.284 3.235 35.689 84.932 92.231 || 33.874 -> 270.990 MByte/s log_avg of all random - random, method 0 : 0.017 0.255 2.965 33.834 82.315 96.998 || 33.026 -> 264.211 MByte/s - random, method 1 : 0.004 0.066 0.930 12.075 49.044 53.915 || 18.089 -> 144.713 MByte/s - random, method 2 : 0.019 0.286 3.258 35.693 85.717 88.625 || 33.234 -> 265.873 MByte/s log_avg(ring,random) - average, method 0 : 0.016 0.255 2.968 33.824 83.762 97.757 || 33.556 -> 268.452 MByte/s - average, method 1 : 0.004 0.062 0.888 11.651 49.658 55.144 || 18.303 -> 146.425 MByte/s - average, method 2 : 0.019 0.285 3.246 35.691 85.323 90.410 || 33.552 -> 268.419 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 0.150 0.019 0.019 0.019 0.016 0.004 0.019 2 0.286 0.036 0.036 0.036 0.032 0.008 0.036 4 0.590 0.074 0.074 0.074 0.065 0.016 0.074 8 1.199 0.150 0.149 0.150 0.131 0.032 0.150 16 2.283 0.285 0.284 0.286 0.255 0.062 0.285 32 4.681 0.585 0.583 0.587 0.507 0.125 0.585 64 9.363 1.170 1.164 1.177 1.013 0.251 1.170 128 17.465 2.183 2.179 2.187 1.908 0.492 2.183 256 25.969 3.246 3.235 3.258 2.968 0.888 3.246 512 49.040 6.130 6.093 6.167 5.641 1.710 6.130 1024 83.917 10.490 10.437 10.542 9.938 3.149 10.490 2048 179.880 22.485 22.454 22.516 20.913 6.529 22.485 4096 285.527 35.691 35.689 35.693 33.824 11.651 35.691 8933 368.237 46.030 46.119 45.940 45.152 17.801 46.030 19484 420.913 52.614 53.750 51.502 46.142 27.394 52.614 42495 591.884 73.986 75.518 72.484 68.164 40.603 73.954 92682 696.317 87.040 88.383 85.717 83.762 49.658 85.323 202141 741.039 92.630 93.605 91.665 92.627 53.869 89.336 440872 776.045 97.006 98.928 95.120 96.759 57.089 92.808 961548 775.032 96.879 99.315 94.503 96.682 57.561 91.263 2097152 782.904 97.863 98.736 96.998 97.757 55.144 90.410 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 92682 2097152 -> average -> accumulated p00 ring-4*2fix : 0.018 0.270 3.152 34.967 103.713 107.773 -> 39.360 -> 314.881 MByte/s p01 ring-2*4fix : 0.019 0.288 3.240 35.875 86.179 102.225 -> 34.852 -> 278.817 MByte/s p02 ring-1*8fix : 0.019 0.288 3.260 35.822 84.588 100.764 -> 34.223 -> 273.784 MByte/s p03 ring-1*8fix : 0.019 0.287 3.256 35.866 85.678 94.707 -> 34.098 -> 272.782 MByte/s p04 ring-1*8fix : 0.019 0.287 3.250 35.842 85.820 93.610 -> 34.232 -> 273.856 MByte/s p05 ring-1*8fix : 0.019 0.287 3.253 35.771 85.744 94.138 -> 34.147 -> 273.178 MByte/s p06 random-cyc-1dim : 0.019 0.286 3.277 35.672 84.971 101.744 -> 34.096 -> 272.764 MByte/s p07 random-cyc-1dim : 0.019 0.288 3.279 35.865 85.428 94.517 -> 34.033 -> 272.262 MByte/s p08 random-cyc-1dim : 0.019 0.289 3.249 35.930 84.868 93.537 -> 33.829 -> 270.631 MByte/s p09 random-cyc-1dim : 0.019 0.287 3.272 35.584 85.118 99.499 -> 34.370 -> 274.961 MByte/s p10 random-cyc-1dim : 0.019 0.287 3.273 35.799 85.932 96.828 -> 34.055 -> 272.441 MByte/s p11 random-cyc-1dim : 0.019 0.287 3.249 35.859 86.748 94.954 -> 34.043 -> 272.340 MByte/s p12 random-cyc-1dim : 0.019 0.287 3.256 35.780 87.744 101.486 -> 34.522 -> 276.177 MByte/s p13 random-cyc-1dim : 0.019 0.288 3.270 35.746 84.326 97.749 -> 34.105 -> 272.842 MByte/s p14 random-cyc-1dim : 0.019 0.278 3.205 35.231 84.923 93.870 -> 34.226 -> 273.807 MByte/s p15 random-cyc-1dim : 0.019 0.285 3.246 35.466 87.174 96.215 -> 34.016 -> 272.126 MByte/s -- additional patterns that are not used to compute b_eff: p16 worst-cyc-1dim : 0.019 0.281 3.255 35.509 85.209 101.410 -> 34.559 -> 276.471 MByte/s p17 best bi-section : 0.018 0.280 2.590 30.250 100.870 111.521 -> 37.373 -> 298.981 MByte/s p18 worst bi-section : 0.018 0.273 2.598 30.219 98.445 112.129 -> 37.271 -> 298.165 MByte/s p19 acyclic-1dim-all : 0.017 0.253 2.856 31.802 75.048 80.284 -> 30.187 -> 241.492 MByte/s p20 acyclic-2dim-all : 0.015 0.229 2.572 28.718 77.652 86.757 -> 30.254 -> 242.036 MByte/s p21 acyclic-3dim-all : 0.018 0.288 2.722 31.048 84.403 105.342 -> 34.545 -> 276.364 MByte/s p22 cyclic-1dim-all : 0.018 0.298 3.280 35.219 85.387 95.708 -> 34.584 -> 276.676 MByte/s p23 cyclic-2dim-all : 0.019 0.300 3.344 35.627 85.753 103.664 -> 35.677 -> 285.412 MByte/s p24 cyclic-3dim-all : 0.019 0.303 3.364 35.787 103.183 106.706 -> 39.401 -> 315.207 MByte/s log_avg of all rings : 0.019 0.284 3.235 35.689 88.383 98.736 || 35.104 -> 280.830 MByte/s log_avg of all random : 0.019 0.286 3.258 35.693 85.717 96.998 || 34.129 -> 273.031 MByte/s log_avg(ring,random) : 0.019 0.285 3.246 35.691 87.040 97.863 || 34.613 -> 276.903 MByte/s SECTION-BY-PATTERN-MSGLNG-END SECTION-BEFF-BEGIN The effective bandwidth is b_eff = 276.903 MByte/s on 8 processes ( = 34.613 MByte/s * 8 processes) system parameters : 8 nodes, 256 MB/node system name: HI-UX/MPP hostname : hitachi OS release : 02-03 OS version : 0 machine : SR2201 SECTION-BEFF-END b_eff = 276.903 MB/s = 34.613 * 8 PEs with 256 MB/PE on HI-UX/MPP hitachi 02-03 0 SR2201