b_eff = 80.086 MB/s = 40.043 * 2 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 = 2 1-dim-paterns: size = 2 2-dim-paterns: size = 2 * 1 3-dim-paterns: size = 2 * 1 * 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 (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 = 140.004 sec sum of max elapsed time per entries above = 139.888 sec difference = 0.116 sec = 0.1% The difference is less than 5 % SECTION-LOOP-END SECTION-PATTERN-BEGIN Pattern parameters: p00 ring-1*2fix => 1 sendrecv_calls with 2 messages, i.e. msgs/used node, all nodes are used p01 ring-1*2fix => 1 sendrecv_calls with 2 messages, i.e. msgs/used node, all nodes are used p02 ring-1*2fix => 1 sendrecv_calls with 2 messages, i.e. msgs/used node, all nodes are used p03 ring-1*2fix => 1 sendrecv_calls with 2 messages, i.e. msgs/used node, all nodes are used p04 ring-1*2fix => 1 sendrecv_calls with 2 messages, i.e. msgs/used node, all nodes are used p05 ring-1*2fix => 1 sendrecv_calls with 2 messages, i.e. msgs/used node, all nodes are used p06 random-cyc-1dim => 1 sendrecv_calls with 2 messages, i.e. msgs/used node, all nodes are used p07 random-cyc-1dim => 1 sendrecv_calls with 2 messages, i.e. msgs/used node, all nodes are used p08 random-cyc-1dim => 1 sendrecv_calls with 2 messages, i.e. msgs/used node, all nodes are used p09 random-cyc-1dim => 1 sendrecv_calls with 2 messages, i.e. msgs/used node, all nodes are used p10 random-cyc-1dim => 1 sendrecv_calls with 2 messages, i.e. msgs/used node, all nodes are used p11 random-cyc-1dim => 1 sendrecv_calls with 2 messages, i.e. msgs/used node, all nodes are used p12 random-cyc-1dim => 1 sendrecv_calls with 2 messages, i.e. msgs/used node, all nodes are used p13 random-cyc-1dim => 1 sendrecv_calls with 2 messages, i.e. msgs/used node, all nodes are used p14 random-cyc-1dim => 1 sendrecv_calls with 2 messages, i.e. msgs/used node, all nodes are used p15 random-cyc-1dim => 1 sendrecv_calls with 2 messages, i.e. msgs/used node, all nodes are used p16 worst-cyc-1dim => 1 sendrecv_calls with 2 messages, i.e. msgs/used node, all nodes are used p17 best bi-section => 2 sendrecv_calls with 2 messages, i.e. msgs/used node, all nodes are used p18 worst bi-section => 2 sendrecv_calls with 2 messages, i.e. msgs/used node, all nodes are used p19 acyclic-1dim-all => 2 sendrecv_calls with 2 messages, i.e. msgs/used node, all nodes are used p20 acyclic-2dim-all => 2 sendrecv_calls with 2 messages, i.e. msgs/used node, all nodes are used p21 acyclic-3dim-all => 2 sendrecv_calls with 2 messages, i.e. msgs/used node, all nodes are used p22 cyclic-1dim-all => 1 sendrecv_calls with 2 messages, i.e. msgs/used node, all nodes are used p23 cyclic-2dim-all => 1 sendrecv_calls with 2 messages, i.e. msgs/used node, all nodes are used p24 cyclic-3dim-all => 1 sendrecv_calls with 2 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-1*2fix : 39.777 36.280 37.187 -> 39.777 -> 79.554 MByte/s p01 ring-1*2fix : 39.908 36.448 36.616 -> 39.908 -> 79.816 MByte/s p02 ring-1*2fix : 39.682 36.063 36.046 -> 39.682 -> 79.364 MByte/s p03 ring-1*2fix : 39.691 36.097 36.746 -> 39.691 -> 79.381 MByte/s p04 ring-1*2fix : 39.647 36.082 36.560 -> 39.647 -> 79.294 MByte/s p05 ring-1*2fix : 39.668 36.050 35.864 -> 39.668 -> 79.337 MByte/s p06 random-cyc-1dim : 39.763 36.130 36.677 -> 39.763 -> 79.525 MByte/s p07 random-cyc-1dim : 39.799 36.068 36.250 -> 39.799 -> 79.598 MByte/s p08 random-cyc-1dim : 39.830 36.026 35.409 -> 39.830 -> 79.660 MByte/s p09 random-cyc-1dim : 39.756 36.064 36.391 -> 39.756 -> 79.513 MByte/s p10 random-cyc-1dim : 39.776 36.089 36.302 -> 39.776 -> 79.552 MByte/s p11 random-cyc-1dim : 39.679 36.030 35.708 -> 39.679 -> 79.357 MByte/s p12 random-cyc-1dim : 39.697 36.058 35.695 -> 39.697 -> 79.393 MByte/s p13 random-cyc-1dim : 39.632 35.977 36.423 -> 39.632 -> 79.263 MByte/s p14 random-cyc-1dim : 39.621 36.050 36.543 -> 39.621 -> 79.242 MByte/s p15 random-cyc-1dim : 39.615 35.982 35.615 -> 39.615 -> 79.229 MByte/s -- additional patterns that are not used to compute b_eff: p16 worst-cyc-1dim : 39.673 36.296 36.449 -> 39.673 -> 79.347 MByte/s p17 best bi-section : 32.368 36.596 38.663 -> 38.663 -> 77.327 MByte/s p18 worst bi-section : 32.668 36.529 38.399 -> 38.399 -> 76.799 MByte/s p19 acyclic-1dim-all : 32.682 36.739 38.370 -> 38.370 -> 76.741 MByte/s p20 acyclic-2dim-all : 32.703 36.480 38.585 -> 38.585 -> 77.170 MByte/s p21 acyclic-3dim-all : 32.706 36.495 38.966 -> 38.966 -> 77.932 MByte/s p22 cyclic-1dim-all : 40.156 36.367 35.508 -> 40.156 -> 80.312 MByte/s p23 cyclic-2dim-all : 39.777 35.955 36.105 -> 39.777 -> 79.554 MByte/s p24 cyclic-3dim-all : 39.652 36.034 35.621 -> 39.652 -> 79.304 MByte/s log_avg of all rings : 39.729 36.170 36.500 || 39.729 -> 79.458 MByte/s log_avg of all random : 39.717 36.047 36.099 || 39.717 -> 79.433 MByte/s log_avg(ring,random) : 39.723 36.108 36.299 ||( 39.723 -> 79.445)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-1*2fix : 39.632 39.763 39.999 -> 39.999 -> 79.999 MByte/s p01 ring-1*2fix : 40.057 40.187 39.861 -> 40.187 -> 80.373 MByte/s p02 ring-1*2fix : 39.807 39.820 39.825 -> 39.825 -> 79.650 MByte/s p03 ring-1*2fix : 39.716 39.938 39.866 -> 39.938 -> 79.876 MByte/s p04 ring-1*2fix : 39.642 39.836 39.750 -> 39.836 -> 79.671 MByte/s p05 ring-1*2fix : 39.702 39.922 39.689 -> 39.922 -> 79.844 MByte/s p06 random-cyc-1dim : 39.781 39.973 39.913 -> 39.973 -> 79.946 MByte/s p07 random-cyc-1dim : 39.501 39.763 39.923 -> 39.923 -> 79.846 MByte/s p08 random-cyc-1dim : 39.702 39.903 40.073 -> 40.073 -> 80.145 MByte/s p09 random-cyc-1dim : 39.917 39.865 39.581 -> 39.917 -> 79.834 MByte/s p10 random-cyc-1dim : 39.861 39.833 39.978 -> 39.978 -> 79.957 MByte/s p11 random-cyc-1dim : 39.931 39.768 39.773 -> 39.931 -> 79.862 MByte/s p12 random-cyc-1dim : 39.632 39.916 39.811 -> 39.916 -> 79.833 MByte/s p13 random-cyc-1dim : 39.731 39.714 39.879 -> 39.879 -> 79.757 MByte/s p14 random-cyc-1dim : 39.840 39.813 39.639 -> 39.840 -> 79.681 MByte/s p15 random-cyc-1dim : 39.723 39.708 39.791 -> 39.791 -> 79.581 MByte/s -- additional patterns that are not used to compute b_eff: p16 worst-cyc-1dim : 40.013 39.930 39.800 -> 40.013 -> 80.026 MByte/s p17 best bi-section : 38.342 38.747 38.861 -> 38.861 -> 77.723 MByte/s p18 worst bi-section : 38.376 38.539 38.294 -> 38.539 -> 77.077 MByte/s p19 acyclic-1dim-all : 38.342 38.935 38.273 -> 38.935 -> 77.870 MByte/s p20 acyclic-2dim-all : 38.556 38.726 38.224 -> 38.726 -> 77.452 MByte/s p21 acyclic-3dim-all : 38.640 38.595 38.566 -> 38.640 -> 77.280 MByte/s p22 cyclic-1dim-all : 39.987 40.308 40.166 -> 40.308 -> 80.617 MByte/s p23 cyclic-2dim-all : 39.762 39.795 39.951 -> 39.951 -> 79.902 MByte/s p24 cyclic-3dim-all : 39.627 39.923 39.922 -> 39.923 -> 79.846 MByte/s log_avg of all rings : 39.759 39.911 39.831 || 39.951 -> 79.902 MByte/s log_avg of all random : 39.762 39.826 39.836 || 39.922 -> 79.844 MByte/s log_avg(ring,random) : 39.760 39.868 39.834 ||( 39.936 -> 79.873)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-1*2fix p00 method 0 : 0.017 0.272 3.122 34.002 104.291 113.066 -> 39.777 -> 79.554 MByte/s p00 method 1 : 0.008 0.133 1.775 22.416 97.476 113.741 -> 36.280 -> 72.560 MByte/s p00 method 2 : 0.019 0.296 3.352 35.425 99.707 100.093 -> 37.187 -> 74.374 MByte/s p01 ring-1*2fix p01 method 0 : 0.017 0.273 3.118 33.537 104.015 115.609 -> 39.908 -> 79.816 MByte/s p01 method 1 : 0.008 0.133 1.777 22.296 97.327 115.973 -> 36.448 -> 72.896 MByte/s p01 method 2 : 0.019 0.295 3.350 35.264 85.379 100.907 -> 36.616 -> 73.232 MByte/s p02 ring-1*2fix p02 method 0 : 0.017 0.270 3.111 34.548 104.463 109.414 -> 39.682 -> 79.364 MByte/s p02 method 1 : 0.008 0.133 1.779 22.486 97.327 109.096 -> 36.063 -> 72.127 MByte/s p02 method 2 : 0.019 0.297 3.347 35.546 93.417 100.970 -> 36.046 -> 72.091 MByte/s p03 ring-1*2fix p03 method 0 : 0.017 0.271 3.117 33.701 103.877 109.523 -> 39.691 -> 79.381 MByte/s p03 method 1 : 0.008 0.133 1.778 22.232 97.359 109.147 -> 36.097 -> 72.194 MByte/s p03 method 2 : 0.019 0.297 3.339 36.054 84.270 100.882 -> 36.746 -> 73.491 MByte/s p04 ring-1*2fix p04 method 0 : 0.017 0.272 3.115 33.931 104.468 108.846 -> 39.647 -> 79.294 MByte/s p04 method 1 : 0.008 0.133 1.778 22.506 97.295 108.943 -> 36.082 -> 72.164 MByte/s p04 method 2 : 0.019 0.297 3.350 35.483 87.797 100.308 -> 36.560 -> 73.120 MByte/s p05 ring-1*2fix p05 method 0 : 0.017 0.272 3.118 34.059 103.967 109.039 -> 39.668 -> 79.337 MByte/s p05 method 1 : 0.008 0.133 1.775 22.294 97.216 109.062 -> 36.050 -> 72.099 MByte/s p05 method 2 : 0.019 0.296 3.346 35.338 82.705 100.386 -> 35.864 -> 71.728 MByte/s p06 random-cyc-1dim p06 method 0 : 0.017 0.271 3.116 33.688 104.419 109.392 -> 39.763 -> 79.525 MByte/s p06 method 1 : 0.008 0.133 1.773 22.496 96.934 109.215 -> 36.130 -> 72.260 MByte/s p06 method 2 : 0.019 0.296 3.347 35.619 84.659 100.150 -> 36.677 -> 73.354 MByte/s p07 random-cyc-1dim p07 method 0 : 0.017 0.270 3.101 34.304 103.814 112.822 -> 39.799 -> 79.598 MByte/s p07 method 1 : 0.008 0.133 1.775 22.462 97.369 108.503 -> 36.068 -> 72.136 MByte/s p07 method 2 : 0.019 0.293 3.328 35.358 83.556 100.218 -> 36.250 -> 72.501 MByte/s p08 random-cyc-1dim p08 method 0 : 0.017 0.272 3.098 33.544 104.291 112.926 -> 39.830 -> 79.660 MByte/s p08 method 1 : 0.008 0.133 1.774 22.448 97.040 109.363 -> 36.026 -> 72.051 MByte/s p08 method 2 : 0.019 0.294 3.351 35.573 81.267 100.626 -> 35.409 -> 70.817 MByte/s p09 random-cyc-1dim p09 method 0 : 0.017 0.273 3.112 33.777 103.729 111.622 -> 39.756 -> 79.513 MByte/s p09 method 1 : 0.008 0.134 1.776 22.327 97.407 109.443 -> 36.064 -> 72.128 MByte/s p09 method 2 : 0.019 0.293 3.342 35.228 84.396 100.155 -> 36.391 -> 72.782 MByte/s p10 random-cyc-1dim p10 method 0 : 0.017 0.272 3.115 34.034 104.095 109.427 -> 39.776 -> 79.552 MByte/s p10 method 1 : 0.008 0.133 1.777 22.252 97.359 109.392 -> 36.089 -> 72.178 MByte/s p10 method 2 : 0.019 0.295 3.338 35.586 81.977 100.386 -> 36.302 -> 72.605 MByte/s p11 random-cyc-1dim p11 method 0 : 0.017 0.271 3.104 33.607 103.861 109.249 -> 39.679 -> 79.357 MByte/s p11 method 1 : 0.008 0.133 1.777 22.357 97.518 108.914 -> 36.030 -> 72.060 MByte/s p11 method 2 : 0.019 0.295 3.342 35.659 82.933 100.103 -> 35.708 -> 71.416 MByte/s p12 random-cyc-1dim p12 method 0 : 0.017 0.272 3.110 33.584 103.739 109.039 -> 39.697 -> 79.393 MByte/s p12 method 1 : 0.008 0.133 1.777 22.324 96.874 108.824 -> 36.058 -> 72.115 MByte/s p12 method 2 : 0.019 0.293 3.339 35.324 93.029 100.232 -> 35.695 -> 71.391 MByte/s p13 random-cyc-1dim p13 method 0 : 0.017 0.270 3.101 33.633 103.898 109.329 -> 39.632 -> 79.263 MByte/s p13 method 1 : 0.008 0.131 1.752 22.161 97.485 109.141 -> 35.977 -> 71.953 MByte/s p13 method 2 : 0.019 0.294 3.331 35.326 84.497 100.544 -> 36.423 -> 72.847 MByte/s p14 random-cyc-1dim p14 method 0 : 0.017 0.270 3.107 34.309 104.015 109.079 -> 39.621 -> 79.242 MByte/s p14 method 1 : 0.008 0.133 1.770 22.384 97.155 109.512 -> 36.050 -> 72.100 MByte/s p14 method 2 : 0.019 0.299 3.349 36.173 87.137 100.160 -> 36.543 -> 73.085 MByte/s p15 random-cyc-1dim p15 method 0 : 0.017 0.271 3.093 33.833 104.275 109.489 -> 39.615 -> 79.229 MByte/s p15 method 1 : 0.008 0.132 1.755 22.124 97.374 108.983 -> 35.982 -> 71.963 MByte/s p15 method 2 : 0.018 0.293 3.313 35.411 82.198 100.136 -> 35.615 -> 71.230 MByte/s -- additional patterns that are not used to compute b_eff: p16 worst-cyc-1dim p16 method 0 : 0.017 0.272 3.100 33.543 103.988 109.500 -> 39.673 -> 79.347 MByte/s p16 method 1 : 0.008 0.133 1.769 22.499 97.184 112.981 -> 36.296 -> 72.591 MByte/s p16 method 2 : 0.019 0.295 3.339 35.974 86.034 100.328 -> 36.449 -> 72.898 MByte/s p17 best bi-section p17 method 0 : 0.018 0.274 2.603 28.354 78.869 105.031 -> 32.368 -> 64.735 MByte/s p17 method 1 : 0.008 0.131 1.737 21.628 97.234 118.444 -> 36.596 -> 73.192 MByte/s p17 method 2 : 0.015 0.233 2.786 30.912 102.669 113.274 -> 38.663 -> 77.327 MByte/s p18 worst bi-section p18 method 0 : 0.018 0.276 2.608 28.241 78.390 107.091 -> 32.668 -> 65.336 MByte/s p18 method 1 : 0.008 0.131 1.735 21.734 97.323 113.920 -> 36.529 -> 73.058 MByte/s p18 method 2 : 0.015 0.234 2.807 31.145 98.251 113.182 -> 38.399 -> 76.799 MByte/s p19 acyclic-1dim-all p19 method 0 : 0.018 0.277 2.597 28.054 78.702 107.069 -> 32.682 -> 65.364 MByte/s p19 method 1 : 0.008 0.131 1.735 21.761 96.713 118.216 -> 36.739 -> 73.477 MByte/s p19 method 2 : 0.015 0.234 2.793 30.825 99.149 113.433 -> 38.370 -> 76.741 MByte/s p20 acyclic-2dim-all p20 method 0 : 0.018 0.276 2.595 28.233 79.117 107.014 -> 32.703 -> 65.407 MByte/s p20 method 1 : 0.008 0.131 1.742 21.795 96.962 114.075 -> 36.480 -> 72.960 MByte/s p20 method 2 : 0.015 0.233 2.792 30.885 99.188 114.088 -> 38.585 -> 77.170 MByte/s p21 acyclic-3dim-all p21 method 0 : 0.019 0.277 2.594 28.086 78.535 107.173 -> 32.706 -> 65.412 MByte/s p21 method 1 : 0.008 0.131 1.737 21.716 97.058 114.082 -> 36.495 -> 72.990 MByte/s p21 method 2 : 0.015 0.233 2.786 30.909 98.067 117.409 -> 38.966 -> 77.932 MByte/s p22 cyclic-1dim-all p22 method 0 : 0.017 0.271 3.118 34.016 104.752 113.673 -> 40.156 -> 80.312 MByte/s p22 method 1 : 0.008 0.133 1.768 22.421 98.227 110.563 -> 36.367 -> 72.733 MByte/s p22 method 2 : 0.019 0.295 3.337 35.293 82.128 100.641 -> 35.508 -> 71.015 MByte/s p23 cyclic-2dim-all p23 method 0 : 0.017 0.267 3.075 33.558 104.062 111.178 -> 39.777 -> 79.554 MByte/s p23 method 1 : 0.008 0.131 1.760 22.116 97.160 109.193 -> 35.955 -> 71.910 MByte/s p23 method 2 : 0.018 0.290 3.303 34.925 81.875 99.902 -> 36.105 -> 72.210 MByte/s p24 cyclic-3dim-all p24 method 0 : 0.017 0.269 3.085 33.437 104.222 108.061 -> 39.652 -> 79.304 MByte/s p24 method 1 : 0.008 0.133 1.766 22.216 96.861 109.581 -> 36.034 -> 72.067 MByte/s p24 method 2 : 0.019 0.293 3.319 34.920 80.833 100.242 -> 35.621 -> 71.242 MByte/s log_avg of all rings - ring, method 0 : 0.017 0.272 3.117 33.961 104.180 110.888 || 39.729 -> 79.458 MByte/s - ring, method 1 : 0.008 0.133 1.777 22.371 97.333 110.959 || 36.170 -> 72.339 MByte/s - ring, method 2 : 0.019 0.296 3.347 35.517 88.688 100.591 || 36.500 -> 73.001 MByte/s log_avg of all random - random, method 0 : 0.017 0.271 3.106 33.830 104.014 110.227 || 39.717 -> 79.433 MByte/s - random, method 1 : 0.008 0.133 1.771 22.333 97.251 109.129 || 36.047 -> 72.095 MByte/s - random, method 2 : 0.019 0.294 3.338 35.525 84.505 100.271 || 36.099 -> 72.198 MByte/s log_avg(ring,random) - average, method 0 : 0.017 0.271 3.111 33.896 104.097 110.557 || 39.723 -> 79.445 MByte/s - average, method 1 : 0.008 0.133 1.774 22.352 97.292 110.040 || 36.108 -> 72.217 MByte/s - average, method 2 : 0.019 0.295 3.343 35.521 86.571 100.431 || 36.299 -> 72.598 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 0.038 0.019 0.019 0.019 0.017 0.008 0.019 2 0.075 0.037 0.037 0.037 0.034 0.017 0.037 4 0.148 0.074 0.074 0.074 0.068 0.033 0.074 8 0.296 0.148 0.148 0.148 0.137 0.067 0.148 16 0.591 0.295 0.296 0.294 0.271 0.133 0.295 32 1.174 0.587 0.588 0.586 0.538 0.264 0.587 64 2.320 1.160 1.162 1.158 1.054 0.523 1.160 128 4.532 2.266 2.269 2.263 2.028 1.026 2.266 256 6.685 3.343 3.347 3.338 3.111 1.774 3.343 512 12.436 6.218 6.209 6.227 5.833 3.410 6.218 1024 22.061 11.031 11.028 11.033 10.393 6.366 11.031 2048 45.802 22.901 22.895 22.908 21.491 13.002 22.901 4096 71.042 35.521 35.517 35.525 33.896 22.352 35.521 8933 93.426 46.713 46.770 46.656 44.834 33.959 46.713 19484 130.722 65.361 65.324 65.398 65.361 54.220 64.061 42495 181.644 90.822 90.935 90.709 90.822 80.282 82.786 92682 208.193 104.097 104.180 104.014 104.097 97.292 86.571 202141 218.378 109.189 109.027 109.351 109.189 105.752 93.433 440872 229.257 114.629 114.331 114.927 114.629 112.904 100.975 961548 231.584 115.792 115.829 115.755 115.792 114.831 103.118 2097152 221.346 110.673 111.076 110.271 110.557 110.040 100.431 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-1*2fix : 0.019 0.296 3.352 35.425 104.291 113.741 -> 40.096 -> 80.191 MByte/s p01 ring-1*2fix : 0.019 0.295 3.350 35.264 104.015 115.973 -> 40.263 -> 80.527 MByte/s p02 ring-1*2fix : 0.019 0.297 3.347 35.546 104.463 109.414 -> 39.979 -> 79.958 MByte/s p03 ring-1*2fix : 0.019 0.297 3.339 36.054 103.877 109.523 -> 40.036 -> 80.072 MByte/s p04 ring-1*2fix : 0.019 0.297 3.350 35.483 104.468 108.943 -> 39.965 -> 79.929 MByte/s p05 ring-1*2fix : 0.019 0.296 3.346 35.338 103.967 109.062 -> 39.976 -> 79.952 MByte/s p06 random-cyc-1dim : 0.019 0.296 3.347 35.619 104.419 109.392 -> 40.082 -> 80.164 MByte/s p07 random-cyc-1dim : 0.019 0.293 3.328 35.358 103.814 112.822 -> 40.068 -> 80.137 MByte/s p08 random-cyc-1dim : 0.019 0.294 3.351 35.573 104.291 112.926 -> 40.177 -> 80.354 MByte/s p09 random-cyc-1dim : 0.019 0.293 3.342 35.228 103.729 111.622 -> 40.077 -> 80.155 MByte/s p10 random-cyc-1dim : 0.019 0.295 3.338 35.586 104.095 109.427 -> 40.079 -> 80.158 MByte/s p11 random-cyc-1dim : 0.019 0.295 3.342 35.659 103.861 109.249 -> 40.033 -> 80.067 MByte/s p12 random-cyc-1dim : 0.019 0.293 3.339 35.324 103.739 109.039 -> 40.001 -> 80.001 MByte/s p13 random-cyc-1dim : 0.019 0.294 3.331 35.326 103.898 109.329 -> 39.943 -> 79.887 MByte/s p14 random-cyc-1dim : 0.019 0.299 3.349 36.173 104.015 109.512 -> 39.954 -> 79.909 MByte/s p15 random-cyc-1dim : 0.018 0.293 3.313 35.411 104.275 109.489 -> 39.920 -> 79.841 MByte/s -- additional patterns that are not used to compute b_eff: p16 worst-cyc-1dim : 0.019 0.295 3.339 35.974 103.988 112.981 -> 40.182 -> 80.364 MByte/s p17 best bi-section : 0.018 0.274 2.786 30.912 102.669 118.444 -> 39.081 -> 78.162 MByte/s p18 worst bi-section : 0.018 0.276 2.807 31.145 98.251 113.920 -> 38.669 -> 77.338 MByte/s p19 acyclic-1dim-all : 0.018 0.277 2.793 30.825 99.149 118.216 -> 38.988 -> 77.976 MByte/s p20 acyclic-2dim-all : 0.018 0.276 2.792 30.885 99.188 114.088 -> 38.928 -> 77.857 MByte/s p21 acyclic-3dim-all : 0.019 0.277 2.786 30.909 98.067 117.409 -> 38.998 -> 77.997 MByte/s p22 cyclic-1dim-all : 0.019 0.295 3.337 35.293 104.752 113.673 -> 40.452 -> 80.905 MByte/s p23 cyclic-2dim-all : 0.018 0.290 3.303 34.925 104.062 111.178 -> 40.094 -> 80.189 MByte/s p24 cyclic-3dim-all : 0.019 0.293 3.319 34.920 104.222 109.581 -> 40.037 -> 80.073 MByte/s log_avg of all rings : 0.019 0.296 3.347 35.517 104.180 111.076 || 40.052 -> 80.105 MByte/s log_avg of all random : 0.019 0.294 3.338 35.525 104.014 110.271 || 40.033 -> 80.067 MByte/s log_avg(ring,random) : 0.019 0.295 3.343 35.521 104.097 110.673 || 40.043 -> 80.086 MByte/s SECTION-BY-PATTERN-MSGLNG-END SECTION-BEFF-BEGIN The effective bandwidth is b_eff = 80.086 MByte/s on 2 processes ( = 40.043 MByte/s * 2 processes) system parameters : 2 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 = 80.086 MB/s = 40.043 * 2 PEs with 256 MB/PE on HI-UX/MPP hitachi 02-03 0 SR2201