b_eff = 306.570 MB/s = 153.285 * 2 PEs with 1024 MB/PE on HI-UX/MPP himiko 03-00 ad2b0 SR8000 SECTION-HEAD-BEGIN b_eff.c, Revision 3.3 from Nov. 29, 1999 MEMORY_PER_PROCESSOR = 1024 MBytes [1M = 1024*1024] 1-dim-paterns: size = 2 2-dim-paterns: size = 2 * 1 3-dim-paterns: size = 2 * 1 * 1 Used message lengths: msglng= 1 2 4 8 16 32 64 128 256 512 1024 2048 4096 10624 27554 71468 185364 480774 1246974 3234251 8388608 Used methods: 0=Sndrcv 1=Alltoal 2=non-blk Used patterns: 0=ring-1*2fix 1=ring-1*2fix 2=ring-1*2fix 3=ring-1*2fix 4=ring-1*2fix 5=ring-1*2fix 6=random-cyc-1dim 7=random-cyc-1dim 8=random-cyc-1dim 9=random-cyc-1dim 10=random-cyc-1dim 11=random-cyc-1dim 12=random-cyc-1dim 13=random-cyc-1dim 14=random-cyc-1dim 15=random-cyc-1dim 16=random-cyc-1dim 17=random-cyc-1dim 18=random-cyc-1dim 19=random-cyc-1dim 20=random-cyc-1dim 21=random-cyc-1dim 22=random-cyc-1dim 23=random-cyc-1dim 24=random-cyc-1dim 25=random-cyc-1dim 26=random-cyc-1dim 27=random-cyc-1dim 28=random-cyc-1dim 29=random-cyc-1dim 30=random-cyc-1dim 31=random-cyc-1dim 32=random-cyc-1dim 33=random-cyc-1dim 34=random-cyc-1dim 35=random-cyc-1dim 36=worst-cyc-1dim 37=best bi-section 38=worst bi-section 39=one PingPong Pair 40=acyclic-2dim-all 41=acyclic-3dim-all 42=cyclic-2dim-x 43=cyclic-2dim-all 44=cyclic-3dim-x 45=cyclic-3dim-all SECTION-HEAD-END SECTION-LOOP-BEGIN measurment loop: 35.828 sec SECTION-LOOP-END SECTION-LOOPLNGS-BEGIN method=0 | ------- repetition 0 --------- ------- repetition 1 --------- ------- repetition 2 --------- [sec] (= Sndrcv)| loop total per pattrn&mthd loop total per pattrn&mthd loop total per pattrn&mthd msg length| lng time minimum maximum lng time minimum maximum lng time minimum maximum ----------| ---- ------- ------- --------- ---- ------- ------- --------- ---- ------- ------- --------- 1 300 3.9e-01 7.9e-03 1.4e-02 142 1.9e-01 3.7e-03 6.6e-03 142 1.9e-01 3.7e-03 6.6e-03 2 150 2.0e-01 3.9e-03 7.0e-03 95 1.3e-01 2.5e-03 4.4e-03 95 1.3e-01 2.5e-03 4.4e-03 4 95 1.3e-01 2.5e-03 4.5e-03 95 1.3e-01 2.5e-03 4.5e-03 95 1.3e-01 2.5e-03 4.5e-03 8 94 1.2e-01 2.5e-03 4.4e-03 94 1.2e-01 2.5e-03 4.4e-03 94 1.2e-01 2.5e-03 4.4e-03 16 95 1.3e-01 2.5e-03 4.7e-03 95 1.3e-01 2.5e-03 4.8e-03 95 1.3e-01 2.5e-03 4.8e-03 32 93 1.3e-01 2.5e-03 4.8e-03 93 1.3e-01 2.5e-03 4.8e-03 93 1.3e-01 2.5e-03 4.7e-03 64 91 1.3e-01 2.5e-03 4.8e-03 91 1.3e-01 2.5e-03 4.8e-03 91 1.3e-01 2.5e-03 4.7e-03 128 90 1.3e-01 2.6e-03 5.0e-03 90 1.3e-01 2.6e-03 5.0e-03 90 1.3e-01 2.6e-03 5.0e-03 256 86 1.3e-01 2.6e-03 4.8e-03 86 1.3e-01 2.6e-03 4.9e-03 86 1.3e-01 2.6e-03 4.8e-03 512 83 1.3e-01 2.5e-03 4.8e-03 83 1.3e-01 2.5e-03 4.8e-03 83 1.3e-01 2.5e-03 4.7e-03 1024 82 1.3e-01 2.6e-03 4.9e-03 82 1.3e-01 2.6e-03 4.8e-03 82 1.3e-01 2.6e-03 4.8e-03 2048 80 2.0e-01 4.0e-03 7.0e-03 80 2.0e-01 4.0e-03 7.0e-03 80 2.0e-01 4.0e-03 7.0e-03 4096 49 1.5e-01 3.0e-03 5.1e-03 49 1.5e-01 3.0e-03 5.0e-03 49 1.5e-01 3.0e-03 5.1e-03 10624 31 1.4e-01 2.7e-03 4.9e-03 31 1.4e-01 2.6e-03 4.9e-03 31 1.4e-01 2.7e-03 4.9e-03 27554 22 1.5e-01 2.9e-03 5.6e-03 22 1.5e-01 2.9e-03 5.7e-03 22 1.5e-01 2.9e-03 5.5e-03 71468 14 1.7e-01 3.3e-03 5.7e-03 14 1.7e-01 3.3e-03 5.7e-03 14 1.7e-01 3.3e-03 5.8e-03 185364 8 1.8e-01 3.4e-03 6.6e-03 8 1.7e-01 3.4e-03 6.2e-03 8 1.7e-01 3.4e-03 6.1e-03 480774 4 2.1e-01 4.2e-03 6.0e-03 4 2.1e-01 4.2e-03 6.0e-03 4 2.1e-01 4.2e-03 6.0e-03 1246974 1 1.2e-01 2.5e-03 3.7e-03 1 1.2e-01 2.5e-03 3.5e-03 1 1.2e-01 2.5e-03 3.5e-03 3234251 1 3.1e-01 6.3e-03 9.0e-03 1 3.1e-01 6.3e-03 9.0e-03 1 3.1e-01 6.3e-03 9.0e-03 8388608 1 7.7e-01 1.6e-02 2.1e-02 1 7.7e-01 1.6e-02 2.1e-02 1 7.7e-01 1.6e-02 2.1e-02 method=1 | ------- repetition 0 --------- ------- repetition 1 --------- ------- repetition 2 --------- [sec] (=Alltoal)| loop total per pattrn&mthd loop total per pattrn&mthd loop total per pattrn&mthd msg length| lng time minimum maximum lng time minimum maximum lng time minimum maximum ----------| ---- ------- ------- --------- ---- ------- ------- --------- ---- ------- ------- --------- 1 300 5.7e-01 1.3e-02 1.3e-02 88 1.7e-01 3.7e-03 3.8e-03 88 1.7e-01 3.7e-03 3.8e-03 2 150 2.9e-01 6.3e-03 6.5e-03 59 1.1e-01 2.5e-03 2.6e-03 59 1.1e-01 2.5e-03 2.6e-03 4 75 1.4e-01 3.2e-03 3.3e-03 59 1.1e-01 2.5e-03 2.6e-03 59 1.1e-01 2.5e-03 2.6e-03 8 59 1.1e-01 2.5e-03 2.6e-03 59 1.1e-01 2.5e-03 2.6e-03 59 1.1e-01 2.5e-03 2.6e-03 16 59 1.1e-01 2.5e-03 2.6e-03 59 1.1e-01 2.5e-03 2.6e-03 59 1.1e-01 2.5e-03 2.6e-03 32 58 1.1e-01 2.5e-03 2.6e-03 58 1.1e-01 2.5e-03 2.6e-03 58 1.1e-01 2.5e-03 2.6e-03 64 58 1.1e-01 2.5e-03 2.6e-03 58 1.1e-01 2.5e-03 2.6e-03 58 1.2e-01 2.5e-03 2.6e-03 128 57 1.2e-01 2.5e-03 2.7e-03 57 1.2e-01 2.5e-03 2.7e-03 57 1.2e-01 2.5e-03 2.7e-03 256 56 1.2e-01 2.6e-03 2.7e-03 56 1.2e-01 2.6e-03 2.8e-03 56 1.2e-01 2.6e-03 2.7e-03 512 54 1.1e-01 2.5e-03 2.6e-03 54 1.1e-01 2.5e-03 2.7e-03 54 1.1e-01 2.5e-03 2.6e-03 1024 54 1.2e-01 2.5e-03 2.7e-03 54 1.2e-01 2.5e-03 2.7e-03 54 1.2e-01 2.5e-03 2.7e-03 2048 53 1.6e-01 3.6e-03 3.8e-03 53 1.6e-01 3.6e-03 3.7e-03 53 1.6e-01 3.6e-03 3.8e-03 4096 36 1.3e-01 2.8e-03 3.0e-03 37 1.3e-01 2.9e-03 3.1e-03 37 1.3e-01 2.9e-03 3.1e-03 10624 24 1.1e-01 2.4e-03 2.7e-03 24 1.1e-01 2.4e-03 2.6e-03 24 1.1e-01 2.4e-03 2.6e-03 27554 18 1.3e-01 2.7e-03 3.0e-03 18 1.3e-01 2.7e-03 3.0e-03 18 1.2e-01 2.7e-03 2.9e-03 71468 12 1.4e-01 3.1e-03 3.4e-03 12 1.4e-01 3.1e-03 3.4e-03 12 1.4e-01 3.1e-03 3.5e-03 185364 7 1.5e-01 3.1e-03 3.6e-03 7 1.5e-01 3.1e-03 3.6e-03 7 1.5e-01 3.1e-03 3.7e-03 480774 4 1.9e-01 4.3e-03 4.5e-03 4 1.9e-01 4.3e-03 4.5e-03 4 1.9e-01 4.3e-03 4.5e-03 1246974 1 1.1e-01 2.5e-03 2.8e-03 1 1.1e-01 2.5e-03 3.2e-03 1 1.1e-01 2.5e-03 2.8e-03 3234251 1 2.9e-01 6.3e-03 1.0e-02 1 2.9e-01 6.3e-03 7.1e-03 1 2.9e-01 6.3e-03 7.1e-03 8388608 1 7.3e-01 1.6e-02 1.6e-02 1 7.3e-01 1.6e-02 1.6e-02 1 7.3e-01 1.6e-02 1.6e-02 method=2 | ------- repetition 0 --------- ------- repetition 1 --------- ------- repetition 2 --------- [sec] (=non-blk)| loop total per pattrn&mthd loop total per pattrn&mthd loop total per pattrn&mthd msg length| lng time minimum maximum lng time minimum maximum lng time minimum maximum ----------| ---- ------- ------- --------- ---- ------- ------- --------- ---- ------- ------- --------- 1 300 6.9e-01 1.5e-02 1.5e-02 73 1.7e-01 3.6e-03 3.8e-03 73 1.6e-01 3.6e-03 3.8e-03 2 150 3.4e-01 7.5e-03 7.7e-03 50 1.1e-01 2.5e-03 2.6e-03 50 1.1e-01 2.5e-03 2.6e-03 4 75 1.7e-01 3.8e-03 3.9e-03 50 1.2e-01 2.5e-03 2.7e-03 50 1.1e-01 2.5e-03 2.6e-03 8 49 1.1e-01 2.5e-03 2.6e-03 49 1.1e-01 2.5e-03 2.6e-03 49 1.1e-01 2.5e-03 2.6e-03 16 49 1.2e-01 2.6e-03 2.7e-03 49 1.2e-01 2.6e-03 2.7e-03 49 1.2e-01 2.6e-03 2.7e-03 32 47 1.1e-01 2.5e-03 2.6e-03 47 1.1e-01 2.5e-03 2.6e-03 47 1.1e-01 2.5e-03 2.6e-03 64 46 1.1e-01 2.5e-03 2.6e-03 47 1.1e-01 2.5e-03 2.6e-03 47 1.1e-01 2.5e-03 2.6e-03 128 46 1.2e-01 2.5e-03 2.6e-03 46 1.1e-01 2.5e-03 2.6e-03 46 1.1e-01 2.5e-03 2.6e-03 256 45 1.1e-01 2.4e-03 2.6e-03 46 1.1e-01 2.5e-03 2.7e-03 45 1.1e-01 2.4e-03 2.6e-03 512 46 1.2e-01 2.5e-03 2.7e-03 46 1.2e-01 2.5e-03 2.7e-03 46 1.2e-01 2.5e-03 2.7e-03 1024 45 1.2e-01 2.5e-03 2.7e-03 45 1.1e-01 2.5e-03 2.7e-03 45 1.1e-01 2.5e-03 2.7e-03 2048 45 1.4e-01 3.1e-03 3.4e-03 45 1.4e-01 3.1e-03 3.3e-03 45 1.4e-01 3.1e-03 3.3e-03 4096 35 1.3e-01 2.8e-03 3.1e-03 36 1.3e-01 2.9e-03 3.1e-03 36 1.3e-01 2.9e-03 3.1e-03 10624 23 1.1e-01 2.4e-03 3.1e-03 23 1.2e-01 2.4e-03 3.1e-03 23 1.2e-01 2.4e-03 3.1e-03 27554 18 1.6e-01 2.7e-03 4.0e-03 18 1.6e-01 2.7e-03 4.1e-03 18 1.5e-01 2.7e-03 4.0e-03 71468 12 1.6e-01 3.1e-03 4.5e-03 12 1.6e-01 3.1e-03 4.6e-03 12 1.6e-01 3.1e-03 4.6e-03 185364 7 1.8e-01 3.1e-03 5.2e-03 7 1.8e-01 3.1e-03 5.2e-03 7 1.8e-01 3.1e-03 4.9e-03 480774 4 2.3e-01 4.3e-03 6.0e-03 4 2.3e-01 4.3e-03 7.0e-03 4 2.3e-01 4.3e-03 7.0e-03 1246974 1 1.2e-01 2.5e-03 3.4e-03 1 1.2e-01 2.5e-03 3.4e-03 1 1.2e-01 2.5e-03 3.4e-03 3234251 1 2.9e-01 6.3e-03 6.9e-03 1 2.9e-01 6.3e-03 6.5e-03 1 2.9e-01 6.3e-03 6.5e-03 8388608 1 7.3e-01 1.6e-02 1.6e-02 1 7.3e-01 1.6e-02 1.6e-02 1 7.3e-01 1.6e-02 1.6e-02 SECTION-LOOPLNGS-END SECTION-ELAPSED-BEGIN measurment loop: elapsed time = 35.828 sec sum of max elapsed time per entries above = 35.501 sec difference to elapsed time = 0.327 sec = 0.9% sum based on fastest repetition = 33.289 sec difference to elapsed time = 2.540 sec = 7.1% CAUTION: A difference above is more than 5 %. There may be problems with the MPI implementation or processes weren't on dedicated processors SECTION-ELAPSED-END SECTION-PATTERN-BEGIN Pattern parameters: sendrecv total number messages/ messages/ unused best method calls of messages used node node&call nodes per repetition ---- ----------------- ----- ------------ --------- --------- ------ -------------- p00 ring-1*2fix 1 2 1.00 1.00 0 ( -1 -1 -1 ) p01 ring-1*2fix 1 2 1.00 1.00 0 ( -1 -1 -1 ) p02 ring-1*2fix 1 2 1.00 1.00 0 ( -1 -1 -1 ) p03 ring-1*2fix 1 2 1.00 1.00 0 ( -1 -1 -1 ) p04 ring-1*2fix 1 2 1.00 1.00 0 ( -1 -1 -1 ) p05 ring-1*2fix 1 2 1.00 1.00 0 ( -1 -1 -1 ) p06 random-cyc-1dim 1 2 1.00 1.00 0 ( -1 -1 -1 ) p07 random-cyc-1dim 1 2 1.00 1.00 0 ( -1 -1 -1 ) p08 random-cyc-1dim 1 2 1.00 1.00 0 ( -1 -1 -1 ) p09 random-cyc-1dim 1 2 1.00 1.00 0 ( -1 -1 -1 ) p10 random-cyc-1dim 1 2 1.00 1.00 0 ( -1 -1 -1 ) p11 random-cyc-1dim 1 2 1.00 1.00 0 ( -1 -1 -1 ) p12 random-cyc-1dim 1 2 1.00 1.00 0 ( -1 -1 -1 ) p13 random-cyc-1dim 1 2 1.00 1.00 0 ( -1 -1 -1 ) p14 random-cyc-1dim 1 2 1.00 1.00 0 ( -1 -1 -1 ) p15 random-cyc-1dim 1 2 1.00 1.00 0 ( -1 -1 -1 ) p16 random-cyc-1dim 1 2 1.00 1.00 0 ( -1 -1 -1 ) p17 random-cyc-1dim 1 2 1.00 1.00 0 ( -1 -1 -1 ) p18 random-cyc-1dim 1 2 1.00 1.00 0 ( -1 -1 -1 ) p19 random-cyc-1dim 1 2 1.00 1.00 0 ( -1 -1 -1 ) p20 random-cyc-1dim 1 2 1.00 1.00 0 ( -1 -1 -1 ) p21 random-cyc-1dim 1 2 1.00 1.00 0 ( -1 -1 -1 ) p22 random-cyc-1dim 1 2 1.00 1.00 0 ( -1 -1 -1 ) p23 random-cyc-1dim 1 2 1.00 1.00 0 ( -1 -1 -1 ) p24 random-cyc-1dim 1 2 1.00 1.00 0 ( -1 -1 -1 ) p25 random-cyc-1dim 1 2 1.00 1.00 0 ( -1 -1 -1 ) p26 random-cyc-1dim 1 2 1.00 1.00 0 ( -1 -1 -1 ) p27 random-cyc-1dim 1 2 1.00 1.00 0 ( -1 -1 -1 ) p28 random-cyc-1dim 1 2 1.00 1.00 0 ( -1 -1 -1 ) p29 random-cyc-1dim 1 2 1.00 1.00 0 ( -1 -1 -1 ) p30 random-cyc-1dim 1 2 1.00 1.00 0 ( -1 -1 -1 ) p31 random-cyc-1dim 1 2 1.00 1.00 0 ( -1 -1 -1 ) p32 random-cyc-1dim 1 2 1.00 1.00 0 ( -1 -1 -1 ) p33 random-cyc-1dim 1 2 1.00 1.00 0 ( -1 -1 -1 ) p34 random-cyc-1dim 1 2 1.00 1.00 0 ( -1 -1 -1 ) p35 random-cyc-1dim 1 2 1.00 1.00 0 ( -1 -1 -1 ) p36 worst-cyc-1dim 1 2 1.00 1.00 0 ( -1 -1 -1 ) p37 best bi-section 2 2 1.00 0.50 0 ( -1 -1 -1 ) p38 worst bi-section 2 2 1.00 0.50 0 ( -1 -1 -1 ) p39 one PingPong Pair 2 2 1.00 0.50 0 ( -1 -1 -1 ) p40 acyclic-2dim-all 2 2 1.00 0.50 0 ( -1 -1 -1 ) p41 acyclic-3dim-all 2 2 1.00 0.50 0 ( -1 -1 -1 ) p42 cyclic-2dim-x 1 2 1.00 1.00 0 ( -1 -1 -1 ) p43 cyclic-2dim-all 1 2 1.00 1.00 0 ( -1 -1 -1 ) p44 cyclic-3dim-x 1 2 1.00 1.00 0 ( -1 -1 -1 ) p45 cyclic-3dim-all 1 2 1.00 1.00 0 ( -1 -1 -1 ) 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 : 151.910 146.439 143.846 -> 151.910 -> 303.819 MByte/s p01 ring-1*2fix : 153.493 146.741 144.940 -> 153.493 -> 306.986 MByte/s p02 ring-1*2fix : 153.177 146.888 138.133 -> 153.177 -> 306.354 MByte/s p03 ring-1*2fix : 152.744 146.131 140.299 -> 152.744 -> 305.489 MByte/s p04 ring-1*2fix : 153.391 145.975 140.618 -> 153.391 -> 306.783 MByte/s p05 ring-1*2fix : 153.481 146.195 136.319 -> 153.481 -> 306.961 MByte/s p06 random-cyc-1dim : 153.280 146.548 140.093 -> 153.280 -> 306.560 MByte/s p07 random-cyc-1dim : 152.067 146.597 135.654 -> 152.067 -> 304.134 MByte/s p08 random-cyc-1dim : 152.722 145.008 145.735 -> 152.722 -> 305.444 MByte/s p09 random-cyc-1dim : 150.694 141.522 145.858 -> 150.694 -> 301.387 MByte/s p10 random-cyc-1dim : 152.907 146.200 138.652 -> 152.907 -> 305.815 MByte/s p11 random-cyc-1dim : 148.100 146.649 140.284 -> 148.100 -> 296.200 MByte/s p12 random-cyc-1dim : 153.818 144.324 137.432 -> 153.818 -> 307.637 MByte/s p13 random-cyc-1dim : 152.814 145.749 142.136 -> 152.814 -> 305.628 MByte/s p14 random-cyc-1dim : 150.506 146.271 143.546 -> 150.506 -> 301.012 MByte/s p15 random-cyc-1dim : 153.326 145.874 140.801 -> 153.326 -> 306.651 MByte/s p16 random-cyc-1dim : 153.936 144.768 143.341 -> 153.936 -> 307.872 MByte/s p17 random-cyc-1dim : 152.993 145.271 145.247 -> 152.993 -> 305.987 MByte/s p18 random-cyc-1dim : 152.201 146.215 145.628 -> 152.201 -> 304.403 MByte/s p19 random-cyc-1dim : 150.401 146.788 142.880 -> 150.401 -> 300.801 MByte/s p20 random-cyc-1dim : 153.555 146.408 137.255 -> 153.555 -> 307.109 MByte/s p21 random-cyc-1dim : 153.550 146.936 137.203 -> 153.550 -> 307.100 MByte/s p22 random-cyc-1dim : 153.587 145.190 141.379 -> 153.587 -> 307.175 MByte/s p23 random-cyc-1dim : 153.364 144.492 145.782 -> 153.364 -> 306.729 MByte/s p24 random-cyc-1dim : 150.536 145.173 137.969 -> 150.536 -> 301.071 MByte/s p25 random-cyc-1dim : 153.314 146.923 134.068 -> 153.314 -> 306.628 MByte/s p26 random-cyc-1dim : 147.685 145.379 142.368 -> 147.685 -> 295.369 MByte/s p27 random-cyc-1dim : 153.251 146.688 141.652 -> 153.251 -> 306.501 MByte/s p28 random-cyc-1dim : 153.031 146.371 141.723 -> 153.031 -> 306.061 MByte/s p29 random-cyc-1dim : 151.264 146.333 141.146 -> 151.264 -> 302.527 MByte/s p30 random-cyc-1dim : 153.796 147.185 138.287 -> 153.796 -> 307.593 MByte/s p31 random-cyc-1dim : 150.251 146.047 138.619 -> 150.251 -> 300.503 MByte/s p32 random-cyc-1dim : 153.361 146.674 141.522 -> 153.361 -> 306.723 MByte/s p33 random-cyc-1dim : 152.730 146.091 136.724 -> 152.730 -> 305.459 MByte/s p34 random-cyc-1dim : 149.041 146.840 130.680 -> 149.041 -> 298.081 MByte/s p35 random-cyc-1dim : 153.043 146.401 141.816 -> 153.043 -> 306.086 MByte/s -- additional patterns that are not used to compute b_eff: p36 worst-cyc-1dim : 150.126 147.008 141.518 -> 150.126 -> 300.252 MByte/s p37 best bi-section : 104.008 145.361 144.353 -> 145.361 -> 290.722 MByte/s p38 worst bi-section : 102.652 145.948 145.369 -> 145.948 -> 291.897 MByte/s p39 one PingPong Pair : 104.122 0.000 0.000 -> 104.122 -> 208.243 MByte/s p40 acyclic-2dim-all : 103.982 145.635 144.671 -> 145.635 -> 291.269 MByte/s p41 acyclic-3dim-all : 104.164 145.425 145.182 -> 145.425 -> 290.849 MByte/s p42 cyclic-2dim-x : 148.150 146.124 138.044 -> 148.150 -> 296.300 MByte/s p43 cyclic-2dim-all : 152.265 146.890 140.905 -> 152.265 -> 304.529 MByte/s p44 cyclic-3dim-x : 153.301 145.930 135.906 -> 153.301 -> 306.602 MByte/s p45 cyclic-3dim-all : 153.417 143.668 134.368 -> 153.417 -> 306.833 MByte/s log_avg of all rings : 153.032 146.394 140.661 || 153.032 -> 306.063 MByte/s log_avg of all random : 152.161 145.893 140.469 || 152.161 -> 304.322 MByte/s log_avg(ring,random) : 152.596 146.143 140.565 ||(152.596 -> 305.191)MByte/s * size -> accumulated on all pr.: 305.191 292.287 281.129 ||(305.191)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 : 148.928 150.006 152.579 -> 152.579 -> 305.157 MByte/s p01 ring-1*2fix : 150.792 152.890 152.676 -> 152.890 -> 305.780 MByte/s p02 ring-1*2fix : 152.233 151.634 151.750 -> 152.233 -> 304.466 MByte/s p03 ring-1*2fix : 152.325 152.528 152.348 -> 152.528 -> 305.057 MByte/s p04 ring-1*2fix : 152.922 151.887 151.986 -> 152.922 -> 305.843 MByte/s p05 ring-1*2fix : 152.565 152.570 152.621 -> 152.621 -> 305.242 MByte/s p06 random-cyc-1dim : 151.840 152.938 152.130 -> 152.938 -> 305.876 MByte/s p07 random-cyc-1dim : 152.383 152.638 152.148 -> 152.638 -> 305.275 MByte/s p08 random-cyc-1dim : 151.424 152.003 152.655 -> 152.655 -> 305.310 MByte/s p09 random-cyc-1dim : 151.377 151.002 152.092 -> 152.092 -> 304.185 MByte/s p10 random-cyc-1dim : 152.403 152.779 152.067 -> 152.779 -> 305.558 MByte/s p11 random-cyc-1dim : 152.562 152.367 151.612 -> 152.562 -> 305.125 MByte/s p12 random-cyc-1dim : 153.258 152.048 152.616 -> 153.258 -> 306.515 MByte/s p13 random-cyc-1dim : 152.854 151.691 152.563 -> 152.854 -> 305.708 MByte/s p14 random-cyc-1dim : 151.655 152.570 152.805 -> 152.805 -> 305.609 MByte/s p15 random-cyc-1dim : 151.377 153.076 152.120 -> 153.076 -> 306.152 MByte/s p16 random-cyc-1dim : 152.623 153.697 152.786 -> 153.697 -> 307.394 MByte/s p17 random-cyc-1dim : 152.101 152.765 152.273 -> 152.765 -> 305.529 MByte/s p18 random-cyc-1dim : 152.637 151.916 152.843 -> 152.843 -> 305.687 MByte/s p19 random-cyc-1dim : 151.667 152.350 151.462 -> 152.350 -> 304.699 MByte/s p20 random-cyc-1dim : 152.549 153.022 151.361 -> 153.022 -> 306.044 MByte/s p21 random-cyc-1dim : 152.765 152.550 152.824 -> 152.824 -> 305.648 MByte/s p22 random-cyc-1dim : 152.708 152.200 152.881 -> 152.881 -> 305.762 MByte/s p23 random-cyc-1dim : 152.316 152.416 152.692 -> 152.692 -> 305.383 MByte/s p24 random-cyc-1dim : 152.094 151.707 152.043 -> 152.094 -> 304.188 MByte/s p25 random-cyc-1dim : 152.068 152.431 152.848 -> 152.848 -> 305.696 MByte/s p26 random-cyc-1dim : 152.012 153.048 152.311 -> 153.048 -> 306.096 MByte/s p27 random-cyc-1dim : 152.888 152.265 152.444 -> 152.888 -> 305.776 MByte/s p28 random-cyc-1dim : 152.760 151.319 151.856 -> 152.760 -> 305.521 MByte/s p29 random-cyc-1dim : 152.533 152.012 152.865 -> 152.865 -> 305.730 MByte/s p30 random-cyc-1dim : 153.292 152.984 153.575 -> 153.575 -> 307.151 MByte/s p31 random-cyc-1dim : 151.298 150.715 151.459 -> 151.459 -> 302.917 MByte/s p32 random-cyc-1dim : 152.800 152.083 152.987 -> 152.987 -> 305.974 MByte/s p33 random-cyc-1dim : 152.148 152.711 152.000 -> 152.711 -> 305.421 MByte/s p34 random-cyc-1dim : 151.621 151.687 152.003 -> 152.003 -> 304.005 MByte/s p35 random-cyc-1dim : 152.730 152.145 152.311 -> 152.730 -> 305.459 MByte/s -- additional patterns that are not used to compute b_eff: p36 worst-cyc-1dim : 151.837 151.676 151.990 -> 151.990 -> 303.980 MByte/s p37 best bi-section : 144.577 145.980 144.157 -> 145.980 -> 291.960 MByte/s p38 worst bi-section : 146.258 146.207 145.489 -> 146.258 -> 292.516 MByte/s p39 one PingPong Pair : 103.342 103.373 103.864 -> 103.864 -> 207.728 MByte/s p40 acyclic-2dim-all : 144.480 144.970 145.735 -> 145.735 -> 291.470 MByte/s p41 acyclic-3dim-all : 146.054 145.455 144.580 -> 146.054 -> 292.108 MByte/s p42 cyclic-2dim-x : 152.477 151.865 151.972 -> 152.477 -> 304.953 MByte/s p43 cyclic-2dim-all : 152.191 151.784 151.687 -> 152.191 -> 304.381 MByte/s p44 cyclic-3dim-x : 152.660 151.297 151.914 -> 152.660 -> 305.320 MByte/s p45 cyclic-3dim-all : 152.869 151.231 151.884 -> 152.869 -> 305.739 MByte/s log_avg of all rings : 151.621 151.916 152.326 || 152.629 -> 305.257 MByte/s log_avg of all random : 152.290 152.303 152.354 || 152.756 -> 305.512 MByte/s log_avg(ring,random) : 151.955 152.110 152.340 ||(152.692 -> 305.385)MByte/s * size -> accumulated on all pr.: 303.911 304.219 304.680 ||(305.385)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 185364 8388608 -> average -> accumulated (latency of one sendrecv (or equivalent method) in microsec) p00 ring-1*2fix p00 method 0 =Sndrcv :( 26.190) 0.038 0.585 8.533 65.526 433.344 520.737 -> 151.910 -> 303.819 MByte/s p00 method 1 =Alltoal :( 42.150) 0.024 0.376 5.600 51.583 415.745 518.775 -> 146.439 -> 292.878 MByte/s p00 method 2 =non-blk :( 50.424) 0.020 0.303 4.658 50.395 411.144 519.867 -> 143.846 -> 287.692 MByte/s p01 ring-1*2fix p01 method 0 =Sndrcv :( 26.254) 0.038 0.595 8.346 65.937 429.573 521.551 -> 153.493 -> 306.986 MByte/s p01 method 1 =Alltoal :( 42.317) 0.024 0.369 5.593 51.462 416.413 519.388 -> 146.741 -> 293.482 MByte/s p01 method 2 =non-blk :( 49.617) 0.020 0.303 4.723 50.223 407.130 518.137 -> 144.940 -> 289.880 MByte/s p02 ring-1*2fix p02 method 0 =Sndrcv :( 26.177) 0.038 0.595 8.381 65.955 425.648 520.029 -> 153.177 -> 306.354 MByte/s p02 method 1 =Alltoal :( 42.317) 0.024 0.376 5.478 51.540 416.015 520.903 -> 146.888 -> 293.775 MByte/s p02 method 2 =non-blk :( 49.575) 0.020 0.296 4.758 50.274 255.880 520.999 -> 138.133 -> 276.265 MByte/s p03 ring-1*2fix p03 method 0 =Sndrcv :( 26.443) 0.038 0.596 8.527 65.913 427.462 521.385 -> 152.744 -> 305.489 MByte/s p03 method 1 =Alltoal :( 41.932) 0.024 0.376 5.489 51.611 411.268 519.771 -> 146.131 -> 292.262 MByte/s p03 method 2 =non-blk :( 49.576) 0.020 0.299 4.700 50.309 410.353 520.548 -> 140.299 -> 280.599 MByte/s p04 ring-1*2fix p04 method 0 =Sndrcv :( 26.327) 0.038 0.594 8.510 65.955 429.202 519.771 -> 153.391 -> 306.783 MByte/s p04 method 1 =Alltoal :( 41.852) 0.024 0.376 5.600 51.426 407.511 521.030 -> 145.975 -> 291.951 MByte/s p04 method 2 =non-blk :( 49.699) 0.020 0.302 4.496 50.274 407.526 519.158 -> 140.618 -> 281.237 MByte/s p05 ring-1*2fix p05 method 0 =Sndrcv :( 26.401) 0.038 0.587 8.523 65.955 425.517 520.579 -> 153.481 -> 306.961 MByte/s p05 method 1 =Alltoal :( 41.898) 0.024 0.376 5.602 51.520 408.291 520.060 -> 146.195 -> 292.390 MByte/s p05 method 2 =non-blk :( 50.657) 0.020 0.303 4.704 50.309 259.405 520.741 -> 136.319 -> 272.638 MByte/s p06 random-cyc-1dim p06 method 0 =Sndrcv :( 26.211) 0.038 0.596 8.321 65.613 422.482 519.323 -> 153.280 -> 306.560 MByte/s p06 method 1 =Alltoal :( 41.875) 0.024 0.369 5.600 51.540 413.628 519.192 -> 146.548 -> 293.096 MByte/s p06 method 2 =non-blk :( 50.763) 0.020 0.301 4.733 50.326 287.709 521.096 -> 140.093 -> 280.186 MByte/s p07 random-cyc-1dim p07 method 0 =Sndrcv :( 26.162) 0.038 0.596 8.352 64.618 434.236 520.903 -> 152.067 -> 304.134 MByte/s p07 method 1 =Alltoal :( 42.012) 0.024 0.376 5.478 51.522 414.163 519.802 -> 146.597 -> 293.194 MByte/s p07 method 2 =non-blk :( 50.220) 0.020 0.298 4.743 50.482 272.190 519.323 -> 135.654 -> 271.309 MByte/s p08 random-cyc-1dim p08 method 0 =Sndrcv :( 26.177) 0.038 0.596 8.520 65.656 419.618 520.317 -> 152.722 -> 305.444 MByte/s p08 method 1 =Alltoal :( 42.047) 0.024 0.377 5.491 51.522 406.750 519.645 -> 145.008 -> 290.015 MByte/s p08 method 2 =non-blk :( 49.534) 0.020 0.296 4.649 50.360 412.047 519.288 -> 145.735 -> 291.471 MByte/s p09 random-cyc-1dim p09 method 0 =Sndrcv :( 26.183) 0.038 0.595 8.501 65.718 427.242 519.998 -> 150.694 -> 301.387 MByte/s p09 method 1 =Alltoal :( 42.127) 0.024 0.376 5.593 51.486 356.277 520.610 -> 141.522 -> 283.043 MByte/s p09 method 2 =non-blk :( 49.588) 0.020 0.302 4.570 50.342 409.458 520.934 -> 145.858 -> 291.716 MByte/s p10 random-cyc-1dim p10 method 0 =Sndrcv :( 26.433) 0.038 0.583 8.523 64.723 427.726 518.940 -> 152.907 -> 305.815 MByte/s p10 method 1 =Alltoal :( 41.841) 0.024 0.375 5.604 50.417 416.142 518.875 -> 146.200 -> 292.399 MByte/s p10 method 2 =non-blk :( 49.506) 0.020 0.303 4.696 50.344 410.353 520.321 -> 138.652 -> 277.303 MByte/s p11 random-cyc-1dim p11 method 0 =Sndrcv :( 26.403) 0.038 0.595 8.336 65.674 429.202 520.321 -> 148.100 -> 296.200 MByte/s p11 method 1 =Alltoal :( 41.943) 0.024 0.369 5.580 50.501 416.413 517.848 -> 146.649 -> 293.299 MByte/s p11 method 2 =non-blk :( 49.808) 0.020 0.302 4.720 50.566 300.009 518.359 -> 140.284 -> 280.567 MByte/s p12 random-cyc-1dim p12 method 0 =Sndrcv :( 26.354) 0.038 0.595 8.526 65.957 432.215 521.420 -> 153.818 -> 307.637 MByte/s p12 method 1 =Alltoal :( 41.932) 0.024 0.375 5.461 51.236 369.346 520.806 -> 144.324 -> 288.648 MByte/s p12 method 2 =non-blk :( 50.383) 0.020 0.296 4.739 50.414 412.562 521.192 -> 137.432 -> 274.864 MByte/s p13 random-cyc-1dim p13 method 0 =Sndrcv :( 26.225) 0.038 0.596 8.520 65.975 428.345 519.967 -> 152.814 -> 305.628 MByte/s p13 method 1 =Alltoal :( 41.806) 0.024 0.376 5.482 50.755 400.597 520.544 -> 145.749 -> 291.498 MByte/s p13 method 2 =non-blk :( 50.768) 0.020 0.298 4.638 50.498 408.935 519.998 -> 142.136 -> 284.272 MByte/s p14 random-cyc-1dim p14 method 0 =Sndrcv :( 26.176) 0.038 0.595 8.520 66.043 428.596 518.615 -> 150.506 -> 301.012 MByte/s p14 method 1 =Alltoal :( 41.887) 0.024 0.376 5.602 51.462 417.483 518.584 -> 146.271 -> 292.542 MByte/s p14 method 2 =non-blk :( 50.477) 0.020 0.302 4.559 50.326 411.517 519.419 -> 143.546 -> 287.093 MByte/s p15 random-cyc-1dim p15 method 0 =Sndrcv :( 26.268) 0.038 0.578 8.504 65.913 427.726 520.579 -> 153.326 -> 306.651 MByte/s p15 method 1 =Alltoal :( 42.007) 0.024 0.376 5.600 51.478 417.083 520.999 -> 145.874 -> 291.748 MByte/s p15 method 2 =non-blk :( 49.643) 0.020 0.302 4.735 50.395 306.385 519.933 -> 140.801 -> 281.601 MByte/s p16 random-cyc-1dim p16 method 0 =Sndrcv :( 26.141) 0.038 0.596 8.365 65.975 429.084 521.161 -> 153.936 -> 307.872 MByte/s p16 method 1 =Alltoal :( 42.027) 0.024 0.370 5.596 51.486 368.521 520.417 -> 144.768 -> 289.536 MByte/s p16 method 2 =non-blk :( 49.630) 0.020 0.302 4.698 49.332 411.922 521.065 -> 143.341 -> 286.682 MByte/s p17 random-cyc-1dim p17 method 0 =Sndrcv :( 26.254) 0.038 0.595 8.540 65.934 434.115 518.263 -> 152.993 -> 305.987 MByte/s p17 method 1 =Alltoal :( 42.297) 0.024 0.374 5.432 51.557 391.646 519.737 -> 145.271 -> 290.543 MByte/s p17 method 2 =non-blk :( 49.699) 0.020 0.297 4.739 49.267 411.517 519.001 -> 145.247 -> 290.494 MByte/s p18 random-cyc-1dim p18 method 0 =Sndrcv :( 26.363) 0.038 0.596 8.520 66.087 429.840 520.579 -> 152.201 -> 304.403 MByte/s p18 method 1 =Alltoal :( 41.830) 0.024 0.376 5.489 51.559 413.220 519.514 -> 146.215 -> 292.429 MByte/s p18 method 2 =non-blk :( 49.588) 0.020 0.301 4.743 50.071 410.740 519.353 -> 145.628 -> 291.255 MByte/s p19 random-cyc-1dim p19 method 0 =Sndrcv :( 26.337) 0.038 0.595 8.547 65.846 418.081 521.196 -> 150.401 -> 300.801 MByte/s p19 method 1 =Alltoal :( 41.864) 0.024 0.376 5.587 51.505 415.079 520.903 -> 146.788 -> 293.576 MByte/s p19 method 2 =non-blk :( 49.534) 0.020 0.302 4.645 49.482 411.392 519.192 -> 142.880 -> 285.761 MByte/s p20 random-cyc-1dim p20 method 0 =Sndrcv :( 26.360) 0.038 0.586 8.504 65.697 430.078 520.417 -> 153.555 -> 307.109 MByte/s p20 method 1 =Alltoal :( 41.909) 0.024 0.376 5.596 51.512 409.196 518.553 -> 146.408 -> 292.816 MByte/s p20 method 2 =non-blk :( 50.743) 0.020 0.303 4.712 50.309 345.917 520.387 -> 137.255 -> 274.510 MByte/s p21 random-cyc-1dim p21 method 0 =Sndrcv :( 26.120) 0.038 0.596 8.359 65.826 433.480 519.353 -> 153.550 -> 307.100 MByte/s p21 method 1 =Alltoal :( 41.863) 0.024 0.370 5.593 51.618 417.211 519.802 -> 146.936 -> 293.872 MByte/s p21 method 2 =non-blk :( 50.233) 0.020 0.301 4.733 50.258 409.319 521.196 -> 137.203 -> 274.405 MByte/s p22 random-cyc-1dim p22 method 0 =Sndrcv :( 26.161) 0.038 0.597 8.514 64.556 430.450 521.289 -> 153.587 -> 307.175 MByte/s p22 method 1 =Alltoal :( 42.167) 0.024 0.376 5.487 51.531 403.089 520.610 -> 145.190 -> 290.379 MByte/s p22 method 2 =non-blk :( 50.179) 0.020 0.297 4.731 49.987 411.392 518.970 -> 141.379 -> 282.759 MByte/s p23 random-cyc-1dim p23 method 0 =Sndrcv :( 26.169) 0.038 0.596 8.547 64.514 426.978 520.903 -> 153.364 -> 306.729 MByte/s p23 method 1 =Alltoal :( 42.170) 0.024 0.376 5.600 51.424 369.986 520.383 -> 144.492 -> 288.984 MByte/s p23 method 2 =non-blk :( 49.506) 0.020 0.304 4.729 50.447 411.657 518.489 -> 145.782 -> 291.564 MByte/s p24 random-cyc-1dim p24 method 0 =Sndrcv :( 26.161) 0.038 0.596 8.504 64.955 419.264 519.353 -> 150.536 -> 301.071 MByte/s p24 method 1 =Alltoal :( 42.157) 0.024 0.376 5.596 52.045 409.319 519.032 -> 145.173 -> 290.347 MByte/s p24 method 2 =non-blk :( 49.547) 0.020 0.304 4.640 50.156 292.834 520.194 -> 137.969 -> 275.937 MByte/s p25 random-cyc-1dim p25 method 0 =Sndrcv :( 26.340) 0.038 0.583 8.507 66.326 432.967 521.289 -> 153.314 -> 306.628 MByte/s p25 method 1 =Alltoal :( 42.133) 0.024 0.376 5.589 52.079 415.872 520.160 -> 146.923 -> 293.847 MByte/s p25 method 2 =non-blk :( 50.055) 0.020 0.303 4.749 50.160 269.535 520.675 -> 134.068 -> 268.137 MByte/s p26 random-cyc-1dim p26 method 0 =Sndrcv :( 26.353) 0.038 0.595 8.374 66.570 434.996 520.741 -> 147.685 -> 295.369 MByte/s p26 method 1 =Alltoal :( 41.784) 0.024 0.369 5.576 52.099 395.948 518.875 -> 145.379 -> 290.759 MByte/s p26 method 2 =non-blk :( 49.767) 0.020 0.301 4.725 50.248 350.597 518.328 -> 142.368 -> 284.736 MByte/s p27 random-cyc-1dim p27 method 0 =Sndrcv :( 26.273) 0.038 0.596 8.520 66.413 426.978 519.097 -> 153.251 -> 306.501 MByte/s p27 method 1 =Alltoal :( 41.864) 0.024 0.375 5.461 51.112 412.703 518.871 -> 146.688 -> 293.376 MByte/s p27 method 2 =non-blk :( 50.660) 0.020 0.296 4.758 50.156 412.187 519.806 -> 141.652 -> 283.305 MByte/s p28 random-cyc-1dim p28 method 0 =Sndrcv :( 26.460) 0.038 0.597 8.530 66.589 422.238 520.321 -> 153.031 -> 306.061 MByte/s p28 method 1 =Alltoal :( 41.955) 0.024 0.376 5.604 51.201 409.720 521.455 -> 146.371 -> 292.742 MByte/s p28 method 2 =non-blk :( 50.685) 0.020 0.301 4.673 50.138 410.230 521.323 -> 141.723 -> 283.447 MByte/s p29 random-cyc-1dim p29 method 0 =Sndrcv :( 26.169) 0.038 0.596 8.523 66.415 428.714 519.514 -> 151.264 -> 302.527 MByte/s p29 method 1 =Alltoal :( 42.012) 0.024 0.376 5.600 52.104 416.557 517.908 -> 146.333 -> 292.667 MByte/s p29 method 2 =non-blk :( 49.699) 0.020 0.302 4.714 50.105 315.780 519.192 -> 141.146 -> 282.291 MByte/s p30 random-cyc-1dim p30 method 0 =Sndrcv :( 26.162) 0.038 0.587 8.494 66.392 434.737 519.545 -> 153.796 -> 307.593 MByte/s p30 method 1 =Alltoal :( 42.057) 0.024 0.376 5.593 52.159 417.227 520.675 -> 147.185 -> 294.369 MByte/s p30 method 2 =non-blk :( 50.136) 0.020 0.302 4.694 50.189 284.238 519.706 -> 138.287 -> 276.574 MByte/s p31 random-cyc-1dim p31 method 0 =Sndrcv :( 26.331) 0.038 0.595 8.368 66.238 376.660 520.356 -> 150.251 -> 300.503 MByte/s p31 method 1 =Alltoal :( 41.987) 0.024 0.370 5.585 51.940 402.836 520.999 -> 146.047 -> 292.093 MByte/s p31 method 2 =non-blk :( 49.508) 0.020 0.302 4.750 50.002 411.657 519.001 -> 138.619 -> 277.238 MByte/s p32 random-cyc-1dim p32 method 0 =Sndrcv :( 26.253) 0.038 0.596 8.527 66.326 430.316 521.161 -> 153.361 -> 306.723 MByte/s p32 method 1 =Alltoal :( 41.784) 0.024 0.376 5.472 52.122 416.684 518.519 -> 146.674 -> 293.349 MByte/s p32 method 2 =non-blk :( 49.821) 0.020 0.294 4.696 50.143 349.460 520.190 -> 141.522 -> 283.044 MByte/s p33 random-cyc-1dim p33 method 0 =Sndrcv :( 26.430) 0.038 0.596 8.514 66.326 425.750 520.190 -> 152.730 -> 305.459 MByte/s p33 method 1 =Alltoal :( 42.007) 0.024 0.376 5.596 52.049 412.844 519.837 -> 146.091 -> 292.181 MByte/s p33 method 2 =non-blk :( 49.753) 0.020 0.301 4.731 50.177 287.261 520.837 -> 136.724 -> 273.448 MByte/s p34 random-cyc-1dim p34 method 0 =Sndrcv :( 26.250) 0.038 0.596 8.520 66.285 389.113 519.514 -> 149.041 -> 298.081 MByte/s p34 method 1 =Alltoal :( 41.875) 0.024 0.375 5.606 52.062 415.872 521.034 -> 146.840 -> 293.680 MByte/s p34 method 2 =non-blk :( 50.537) 0.020 0.303 4.716 50.056 286.943 520.837 -> 130.680 -> 261.360 MByte/s p35 random-cyc-1dim p35 method 0 =Sndrcv :( 26.106) 0.038 0.586 8.501 66.305 426.013 519.419 -> 153.043 -> 306.086 MByte/s p35 method 1 =Alltoal :( 42.012) 0.024 0.375 5.602 52.062 401.232 520.452 -> 146.401 -> 292.802 MByte/s p35 method 2 =non-blk :( 50.900) 0.020 0.302 4.743 50.143 411.797 519.580 -> 141.816 -> 283.632 MByte/s -- additional patterns that are not used to compute b_eff: p36 worst-cyc-1dim p36 method 0 =Sndrcv :( 26.357) 0.038 0.596 8.340 66.305 428.330 519.549 -> 150.126 -> 300.252 MByte/s p36 method 1 =Alltoal :( 41.920) 0.024 0.370 5.516 52.099 415.491 518.871 -> 147.008 -> 294.016 MByte/s p36 method 2 =non-blk :( 50.001) 0.020 0.302 4.673 50.265 321.498 520.352 -> 141.518 -> 283.037 MByte/s p37 best bi-section p37 method 0 =Sndrcv :( 22.947) 0.022 0.325 4.591 40.620 248.018 398.640 -> 104.008 -> 208.017 MByte/s p37 method 1 =Alltoal :( 21.250) 0.024 0.362 5.443 51.450 398.266 518.008 -> 145.361 -> 290.722 MByte/s p37 method 2 =non-blk :( 25.020) 0.020 0.304 4.756 49.023 413.895 519.192 -> 144.353 -> 288.706 MByte/s p38 worst bi-section p38 method 0 =Sndrcv :( 23.047) 0.022 0.324 4.602 40.529 247.234 398.678 -> 102.652 -> 205.303 MByte/s p38 method 1 =Alltoal :( 21.324) 0.023 0.363 5.432 51.486 415.348 520.775 -> 145.948 -> 291.897 MByte/s p38 method 2 =non-blk :( 25.513) 0.020 0.304 4.752 49.282 411.268 519.097 -> 145.369 -> 290.739 MByte/s p39 one PingPong Pair p39 method 0 =Sndrcv :( 22.923) 0.022 0.324 4.592 40.686 246.494 399.838 -> 104.122 -> 208.243 MByte/s p39 method 1 =Alltoal :(-------) 0.000 0.000 0.000 0.000 0.000 0.000 -> 0.000 -> 0.000 MByte/s p39 method 2 =non-blk :(-------) 0.000 0.000 0.000 0.000 0.000 0.000 -> 0.000 -> 0.000 MByte/s p40 acyclic-2dim-all p40 method 0 =Sndrcv :( 23.137) 0.022 0.322 4.645 40.269 242.619 399.704 -> 103.982 -> 207.964 MByte/s p40 method 1 =Alltoal :( 21.403) 0.023 0.369 5.459 51.391 415.745 520.256 -> 145.635 -> 291.269 MByte/s p40 method 2 =non-blk :( 25.274) 0.020 0.305 4.756 49.365 411.268 520.063 -> 144.671 -> 289.342 MByte/s p41 acyclic-3dim-all p41 method 0 =Sndrcv :( 23.007) 0.022 0.322 4.628 40.423 246.455 399.648 -> 104.164 -> 208.327 MByte/s p41 method 1 =Alltoal :( 21.380) 0.023 0.369 5.463 50.891 400.597 519.514 -> 145.425 -> 290.849 MByte/s p41 method 2 =non-blk :( 25.027) 0.020 0.306 4.750 49.400 412.171 520.125 -> 145.182 -> 290.363 MByte/s p42 cyclic-2dim-x p42 method 0 =Sndrcv :( 26.440) 0.038 0.587 8.543 66.261 430.450 520.514 -> 148.150 -> 296.300 MByte/s p42 method 1 =Alltoal :( 41.807) 0.024 0.376 5.593 52.062 403.463 518.871 -> 146.124 -> 292.247 MByte/s p42 method 2 =non-blk :( 49.712) 0.020 0.302 4.660 50.138 283.365 519.641 -> 138.044 -> 276.088 MByte/s p43 cyclic-2dim-all p43 method 0 =Sndrcv :( 26.377) 0.038 0.595 8.503 66.457 421.523 520.129 -> 152.265 -> 304.529 MByte/s p43 method 1 =Alltoal :( 41.818) 0.024 0.375 5.589 52.043 416.286 521.196 -> 146.890 -> 293.781 MByte/s p43 method 2 =non-blk :( 49.562) 0.020 0.302 4.735 50.240 404.979 520.964 -> 140.905 -> 281.811 MByte/s p44 cyclic-3dim-x p44 method 0 =Sndrcv :( 26.400) 0.038 0.596 8.523 66.413 426.743 520.448 -> 153.301 -> 306.602 MByte/s p44 method 1 =Alltoal :( 41.852) 0.024 0.370 5.589 52.062 413.487 519.323 -> 145.930 -> 291.860 MByte/s p44 method 2 =non-blk :( 50.835) 0.020 0.301 4.737 50.020 314.403 520.741 -> 135.906 -> 271.812 MByte/s p45 cyclic-3dim-all p45 method 0 =Sndrcv :( 26.161) 0.038 0.596 8.514 66.413 433.722 517.783 -> 153.417 -> 306.833 MByte/s p45 method 1 =Alltoal :( 42.170) 0.024 0.376 5.482 52.099 414.684 519.580 -> 143.668 -> 287.336 MByte/s p45 method 2 =non-blk :( 50.301) 0.020 0.297 4.671 50.122 258.942 518.967 -> 134.368 -> 268.735 MByte/s log_avg of all rings - ring, method 0 = Sndrcv : 0.038 0.592 8.470 65.873 428.449 520.675 || 153.032 -> 306.063 MByte/s - ring, method 1 = Alltoal: 0.024 0.375 5.560 51.524 412.524 519.987 || 146.394 -> 292.789 MByte/s - ring, method 2 = non-blk: 0.020 0.301 4.673 50.298 350.625 519.908 || 140.661 -> 281.321 MByte/s log_avg of all random - random, method 0 = Sndrcv : 0.038 0.593 8.480 65.846 425.054 520.134 || 152.161 -> 304.322 MByte/s - random, method 1 = Alltoal: 0.024 0.374 5.558 51.569 404.393 519.820 || 145.893 -> 291.786 MByte/s - random, method 2 = non-blk: 0.020 0.301 4.703 50.162 359.966 519.905 || 140.469 -> 280.938 MByte/s log_avg(ring,random) - average, method 0 = Sndrcv : 0.038 0.593 8.475 65.860 426.748 520.405 || 152.596 -> 305.191 MByte/s - average, method 1 = Alltoal: 0.024 0.375 5.559 51.547 408.438 519.904 || 146.143 -> 292.287 MByte/s - average, method 2 = non-blk: 0.020 0.301 4.688 50.230 355.265 519.906 || 140.565 -> 281.129 MByte/s * size -> accumulated on all processes: - accumulated, mthd 0 = Sndrcv : 0.076 1.185 16.949 131.719 853.497 1040.809 || 305.191 MByte/s - accumulated, mthd 1 = Alltoal: 0.048 0.749 11.118 103.093 816.877 1039.808 || 292.287 MByte/s - accumulated, mthd 2 = non-blk: 0.040 0.601 9.376 100.459 710.529 1039.813 || 281.129 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: | effective| average rings random method_0 method_1 method_2 | bandwidth| crt,rnd only only Sndrcv Alltoal non-blk 1 0.076 0.038 0.038 0.038 0.038 0.024 0.020 2 0.152 0.076 0.076 0.076 0.076 0.048 0.040 4 0.302 0.151 0.151 0.151 0.151 0.095 0.079 8 0.606 0.303 0.302 0.303 0.303 0.190 0.159 16 1.185 0.593 0.592 0.593 0.593 0.375 0.301 32 2.340 1.170 1.170 1.171 1.170 0.743 0.597 64 4.585 2.292 2.294 2.291 2.292 1.465 1.186 128 8.796 4.398 4.400 4.396 4.398 2.844 2.316 256 16.949 8.475 8.470 8.480 8.475 5.559 4.688 512 33.551 16.775 16.788 16.763 16.775 11.063 9.310 1024 65.393 32.696 32.685 32.708 32.696 21.671 18.392 2048 80.874 40.437 40.357 40.517 40.437 30.232 29.469 4096 131.719 65.860 65.873 65.846 65.860 51.547 50.230 10624 246.275 123.137 122.888 123.388 123.137 104.042 99.997 27554 416.501 208.250 208.401 208.100 208.250 182.471 169.955 71468 593.043 296.521 296.284 296.760 296.521 272.304 254.860 185364 855.709 427.855 428.449 427.261 426.748 408.438 355.265 480774 907.800 453.900 453.212 454.589 449.907 448.373 420.842 1246974 1006.620 503.310 503.969 502.651 499.784 498.399 494.450 3234251 1023.602 511.801 511.857 511.745 505.618 508.926 510.351 8388608 1041.770 520.885 521.074 520.697 520.405 519.904 519.906 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 185364 8388608 -> average -> accumulated (latency of one sendrecv (or equivalent method) in microsec) p00 ring-1*2fix :( 26.190) 0.038 0.585 8.533 65.526 433.344 520.737 -> 153.031 -> 306.061 MByte/s p01 ring-1*2fix :( 26.254) 0.038 0.595 8.346 65.937 429.573 521.551 -> 153.493 -> 306.986 MByte/s p02 ring-1*2fix :( 26.177) 0.038 0.595 8.381 65.955 425.648 520.999 -> 153.223 -> 306.447 MByte/s p03 ring-1*2fix :( 26.443) 0.038 0.596 8.527 65.913 427.462 521.385 -> 153.110 -> 306.221 MByte/s p04 ring-1*2fix :( 26.327) 0.038 0.594 8.510 65.955 429.202 521.030 -> 153.477 -> 306.953 MByte/s p05 ring-1*2fix :( 26.401) 0.038 0.587 8.523 65.955 425.517 520.741 -> 153.488 -> 306.977 MByte/s p06 random-cyc-1dim :( 26.211) 0.038 0.596 8.321 65.613 422.482 521.096 -> 153.365 -> 306.729 MByte/s p07 random-cyc-1dim :( 26.162) 0.038 0.596 8.352 64.618 434.236 520.903 -> 153.128 -> 306.256 MByte/s p08 random-cyc-1dim :( 26.177) 0.038 0.596 8.520 65.656 419.618 520.317 -> 152.822 -> 305.643 MByte/s p09 random-cyc-1dim :( 26.183) 0.038 0.595 8.501 65.718 427.242 520.934 -> 153.286 -> 306.572 MByte/s p10 random-cyc-1dim :( 26.433) 0.038 0.583 8.523 64.723 427.726 520.321 -> 153.261 -> 306.523 MByte/s p11 random-cyc-1dim :( 26.403) 0.038 0.595 8.336 65.674 429.202 520.321 -> 152.986 -> 305.972 MByte/s p12 random-cyc-1dim :( 26.354) 0.038 0.595 8.526 65.957 432.215 521.420 -> 153.869 -> 307.738 MByte/s p13 random-cyc-1dim :( 26.225) 0.038 0.596 8.520 65.975 428.345 520.544 -> 153.179 -> 306.357 MByte/s p14 random-cyc-1dim :( 26.176) 0.038 0.595 8.520 66.043 428.596 519.419 -> 153.049 -> 306.098 MByte/s p15 random-cyc-1dim :( 26.268) 0.038 0.578 8.504 65.913 427.726 520.999 -> 153.353 -> 306.707 MByte/s p16 random-cyc-1dim :( 26.141) 0.038 0.596 8.365 65.975 429.084 521.161 -> 153.936 -> 307.872 MByte/s p17 random-cyc-1dim :( 26.254) 0.038 0.595 8.540 65.934 434.115 519.737 -> 153.267 -> 306.534 MByte/s p18 random-cyc-1dim :( 26.363) 0.038 0.596 8.520 66.087 429.840 520.579 -> 153.229 -> 306.457 MByte/s p19 random-cyc-1dim :( 26.337) 0.038 0.595 8.547 65.846 418.081 521.196 -> 152.818 -> 305.636 MByte/s p20 random-cyc-1dim :( 26.360) 0.038 0.586 8.504 65.697 430.078 520.417 -> 153.555 -> 307.109 MByte/s p21 random-cyc-1dim :( 26.120) 0.038 0.596 8.359 65.826 433.480 521.196 -> 153.837 -> 307.674 MByte/s p22 random-cyc-1dim :( 26.161) 0.038 0.597 8.514 64.556 430.450 521.289 -> 153.618 -> 307.236 MByte/s p23 random-cyc-1dim :( 26.169) 0.038 0.596 8.547 64.514 426.978 520.903 -> 153.364 -> 306.729 MByte/s p24 random-cyc-1dim :( 26.161) 0.038 0.596 8.504 64.955 419.264 520.194 -> 152.990 -> 305.980 MByte/s p25 random-cyc-1dim :( 26.340) 0.038 0.583 8.507 66.326 432.967 521.289 -> 153.314 -> 306.628 MByte/s p26 random-cyc-1dim :( 26.353) 0.038 0.595 8.374 66.570 434.996 520.741 -> 153.550 -> 307.099 MByte/s p27 random-cyc-1dim :( 26.273) 0.038 0.596 8.520 66.413 426.978 519.806 -> 153.284 -> 306.569 MByte/s p28 random-cyc-1dim :( 26.460) 0.038 0.597 8.530 66.589 422.238 521.455 -> 153.085 -> 306.169 MByte/s p29 random-cyc-1dim :( 26.169) 0.038 0.596 8.523 66.415 428.714 519.514 -> 153.673 -> 307.346 MByte/s p30 random-cyc-1dim :( 26.162) 0.038 0.587 8.494 66.392 434.737 520.675 -> 153.850 -> 307.701 MByte/s p31 random-cyc-1dim :( 26.331) 0.038 0.595 8.368 66.238 411.657 520.999 -> 152.102 -> 304.204 MByte/s p32 random-cyc-1dim :( 26.253) 0.038 0.596 8.527 66.326 430.316 521.161 -> 153.619 -> 307.239 MByte/s p33 random-cyc-1dim :( 26.430) 0.038 0.596 8.514 66.326 425.750 520.837 -> 152.934 -> 305.868 MByte/s p34 random-cyc-1dim :( 26.250) 0.038 0.596 8.520 66.285 415.872 521.034 -> 152.605 -> 305.210 MByte/s p35 random-cyc-1dim :( 26.106) 0.038 0.586 8.501 66.305 426.013 520.452 -> 153.092 -> 306.184 MByte/s -- additional patterns that are not used to compute b_eff: p36 worst-cyc-1dim :( 26.357) 0.038 0.596 8.340 66.305 428.330 520.352 -> 152.758 -> 305.516 MByte/s p37 best bi-section :( 21.250) 0.024 0.362 5.443 51.450 413.895 519.192 -> 146.402 -> 292.805 MByte/s p38 worst bi-section :( 21.324) 0.023 0.363 5.432 51.486 415.348 520.775 -> 146.878 -> 293.757 MByte/s p39 one PingPong Pair :( 22.923) 0.022 0.324 4.592 40.686 246.494 399.838 -> 104.122 -> 208.243 MByte/s p40 acyclic-2dim-all :( 21.403) 0.023 0.369 5.459 51.391 415.745 520.256 -> 146.183 -> 292.366 MByte/s p41 acyclic-3dim-all :( 21.380) 0.023 0.369 5.463 50.891 412.171 520.125 -> 146.324 -> 292.648 MByte/s p42 cyclic-2dim-x :( 26.440) 0.038 0.587 8.543 66.261 430.450 520.514 -> 153.066 -> 306.131 MByte/s p43 cyclic-2dim-all :( 26.377) 0.038 0.595 8.503 66.457 421.523 521.196 -> 152.748 -> 305.497 MByte/s p44 cyclic-3dim-x :( 26.400) 0.038 0.596 8.523 66.413 426.743 520.741 -> 153.315 -> 306.629 MByte/s p45 cyclic-3dim-all :( 26.161) 0.038 0.596 8.514 66.413 433.722 519.580 -> 153.502 -> 307.005 MByte/s log_avg of all rings : 0.038 0.592 8.470 65.873 428.449 521.074 || 153.304 -> 306.607 MByte/s log_avg of all random : 0.038 0.593 8.480 65.846 427.261 520.697 || 153.267 -> 306.534 MByte/s log_avg(ring,random) : 0.038 0.593 8.475 65.860 427.855 520.885 || 153.285 -> 306.570 MByte/s * size -> accumulated on all pr.: 0.076 1.185 16.949 131.719 855.709 1041.770 || 306.570 MByte/s SECTION-BY-PATTERN-MSGLNG-END SECTION-BEFF-BEGIN The effective bandwidth is b_eff = 306.570 MByte/s on 2 processes ( = 153.285 MByte/s * 2 processes) Ping-pong latency: 22.923 microsec Ping-pong bandwidth: 799.677 MByte/s at Lmax= 8.000 MByte (MByte/s=1e6 Byte/s) (MByte=2**20 Byte) system parameters : 2 nodes, 1024 MB/node system name : HI-UX/MPP hostname : himiko OS release : 03-00 OS version : ad2b0 machine : SR8000 Date of measurement: Mon Nov 29 17:23:39 1999 Total execution wall clock time = 37 seconds SECTION-BEFF-END b_eff = 306.570 MB/s = 153.285 * 2 PEs with 1024 MB/PE on HI-UX/MPP himiko 03-00 ad2b0 SR8000