#--------------------------------------------------- # PALLAS MPI Benchmark Suite V2.2, MPI-1 part #--------------------------------------------------- # Date : Tue Oct 15 15:27:34 2002 # Machine : i686# System : Linux # Release : 2.4.19 # Version : #1 SMP Tue Oct 8 19:37:08 EDT 2002 # # Minimum message length in bytes: 0 # Maximum message length in bytes: 4194304 # # MPI_Datatype : MPI_BYTE # MPI_Datatype for reductions : MPI_FLOAT # MPI_Op : MPI_SUM # # # List of Benchmarks to run: # PingPong # PingPing # Sendrecv # Exchange # Allreduce # Reduce # Reduce_scatter # Allgather # Allgatherv # Alltoall # Bcast # Barrier #--------------------------------------------------- # Benchmarking PingPong # ( #processes = 2 ) # ( 6 additional processes waiting in MPI_Barrier) #--------------------------------------------------- #bytes #repetitions t[usec] Mbytes/sec 0 1000 8.05 0.00 1 1000 8.14 0.12 2 1000 8.14 0.23 4 1000 8.18 0.47 8 1000 8.29 0.92 16 1000 8.61 1.77 32 1000 8.77 3.48 64 1000 9.03 6.76 128 1000 12.08 10.10 256 1000 14.83 16.46 512 1000 17.89 27.29 1024 1000 24.31 40.17 2048 1000 36.88 52.95 4096 1000 55.67 70.17 8192 1000 95.27 82.00 16384 1000 116.47 134.15 32768 1000 186.97 167.14 65536 640 327.91 190.60 131072 320 609.82 204.98 262144 160 1174.19 212.91 524288 80 2303.11 217.10 1048576 40 4561.19 219.24 2097152 20 8898.06 224.77 4194304 10 17573.37 227.62 #--------------------------------------------------- # Benchmarking PingPing # ( #processes = 2 ) # ( 6 additional processes waiting in MPI_Barrier) #--------------------------------------------------- #bytes #repetitions t[usec] Mbytes/sec 0 1000 12.58 0.00 1 1000 12.78 0.07 2 1000 12.84 0.15 4 1000 12.49 0.31 8 1000 13.21 0.58 16 1000 13.09 1.17 32 1000 13.36 2.28 64 1000 13.41 4.55 128 1000 14.77 8.26 256 1000 19.41 12.58 512 1000 21.69 22.51 1024 1000 28.01 34.87 2048 1000 39.46 49.50 4096 1000 65.57 59.57 8192 1000 108.32 72.12 16384 1000 138.33 112.95 32768 1000 216.11 144.60 65536 640 374.74 166.78 131072 320 694.07 180.10 262144 160 1334.79 187.30 524288 80 2618.32 190.96 1048576 40 5183.59 192.92 2097152 20 10134.70 197.34 4194304 10 20055.55 199.45 #----------------------------------------------------------------------------- # Benchmarking Sendrecv # ( #processes = 2 ) # ( 6 additional processes waiting in MPI_Barrier) #----------------------------------------------------------------------------- #bytes #repetitions t_min[usec] t_max[usec] t_avg[usec] Mbytes/sec 0 1000 13.22 13.22 13.22 0.00 1 1000 13.10 13.10 13.10 0.15 2 1000 13.08 13.10 13.09 0.29 4 1000 13.03 13.13 13.08 0.58 8 1000 13.10 13.14 13.12 1.16 16 1000 13.64 13.68 13.66 2.23 32 1000 13.81 13.82 13.82 4.41 64 1000 13.99 14.05 14.02 8.69 128 1000 15.30 15.32 15.31 15.94 256 1000 19.75 19.76 19.76 24.71 512 1000 21.99 22.04 22.02 44.31 1024 1000 28.60 28.65 28.63 68.17 2048 1000 41.23 41.27 41.25 94.65 4096 1000 65.49 65.51 65.50 119.25 8192 1000 105.73 105.76 105.75 147.74 16384 1000 138.72 138.73 138.72 225.26 32768 1000 216.60 216.61 216.61 288.54 65536 640 374.88 374.97 374.92 333.36 131072 320 694.18 694.72 694.45 359.86 262144 160 1335.50 1335.58 1335.54 374.37 524288 80 2617.22 2618.28 2617.75 381.93 1048576 40 5182.00 5182.25 5182.12 385.93 2097152 20 10133.28 10133.35 10133.31 394.74 4194304 10 20064.95 20069.70 20067.32 398.61 #----------------------------------------------------------------------------- # Benchmarking Sendrecv # ( #processes = 4 ) # ( 4 additional processes waiting in MPI_Barrier) #----------------------------------------------------------------------------- #bytes #repetitions t_min[usec] t_max[usec] t_avg[usec] Mbytes/sec 0 1000 13.79 13.88 13.85 0.00 1 1000 13.89 13.98 13.92 0.14 2 1000 13.84 13.93 13.87 0.27 4 1000 14.08 14.19 14.13 0.54 8 1000 13.93 13.99 13.95 1.09 16 1000 14.02 14.12 14.07 2.16 32 1000 14.19 14.29 14.26 4.27 64 1000 14.38 14.40 14.39 8.48 128 1000 16.75 16.81 16.77 14.52 256 1000 18.35 18.45 18.40 26.47 512 1000 21.74 21.79 21.76 44.83 1024 1000 28.66 28.73 28.69 67.99 2048 1000 42.95 43.08 42.99 90.68 4096 1000 66.06 66.16 66.12 118.08 8192 1000 115.32 115.44 115.37 135.36 16384 1000 192.12 192.17 192.15 162.61 32768 1000 276.40 276.55 276.46 226.00 65536 640 441.01 441.07 441.03 283.40 131072 320 753.15 753.73 753.44 331.68 262144 160 1396.15 1396.84 1396.62 357.95 524288 80 2712.10 2713.68 2713.20 368.50 1048576 40 5366.20 5368.12 5367.14 372.57 2097152 20 10507.55 10516.55 10511.98 380.35 4194304 10 20810.15 20830.65 20820.51 384.05 #----------------------------------------------------------------------------- # Benchmarking Sendrecv # ( #processes = 8 ) #----------------------------------------------------------------------------- #bytes #repetitions t_min[usec] t_max[usec] t_avg[usec] Mbytes/sec 0 1000 14.07 14.14 14.11 0.00 1 1000 14.20 14.33 14.27 0.13 2 1000 14.20 14.30 14.22 0.27 4 1000 14.17 14.28 14.25 0.53 8 1000 14.10 14.15 14.13 1.08 16 1000 14.07 14.18 14.11 2.15 32 1000 14.27 14.48 14.42 4.22 64 1000 14.31 14.37 14.34 8.50 128 1000 17.36 17.46 17.41 13.98 256 1000 19.45 19.62 19.53 24.89 512 1000 22.80 22.95 22.85 42.56 1024 1000 29.13 29.25 29.19 66.77 2048 1000 43.87 43.99 43.93 88.80 4096 1000 65.91 66.07 65.97 118.25 8192 1000 114.22 114.45 114.33 136.52 16384 1000 176.19 176.29 176.21 177.26 32768 1000 272.25 272.40 272.32 229.44 65536 640 456.35 456.69 456.52 273.71 131072 320 823.02 823.89 823.46 303.44 262144 160 1550.72 1551.44 1551.13 322.28 524288 80 3010.62 3012.79 3011.51 331.92 1048576 40 5959.12 5961.44 5960.52 335.49 2097152 20 11693.95 11750.25 11725.12 340.42 4194304 10 23101.60 23412.50 23258.45 341.70 #----------------------------------------------------------------------------- # Benchmarking Exchange # ( #processes = 2 ) # ( 6 additional processes waiting in MPI_Barrier) #----------------------------------------------------------------------------- #bytes #repetitions t_min[usec] t_max[usec] t_avg[usec] Mbytes/sec 0 1000 22.56 22.61 22.58 0.00 1 1000 22.51 22.52 22.51 0.17 2 1000 22.57 22.60 22.58 0.34 4 1000 22.43 22.47 22.45 0.68 8 1000 22.41 22.43 22.42 1.36 16 1000 22.47 22.49 22.48 2.71 32 1000 22.89 22.96 22.93 5.32 64 1000 23.02 23.05 23.03 10.59 128 1000 28.27 28.40 28.34 17.19 256 1000 28.67 28.74 28.71 33.98 512 1000 31.06 31.07 31.06 62.87 1024 1000 37.79 37.83 37.81 103.27 2048 1000 57.92 57.95 57.93 134.82 4096 1000 95.90 95.91 95.91 162.91 8192 1000 165.73 165.84 165.79 188.43 16384 1000 273.45 273.56 273.50 228.47 32768 1000 430.46 430.47 430.47 290.38 65536 640 748.64 748.76 748.70 333.89 131072 320 1383.29 1383.39 1383.34 361.43 262144 160 2661.35 2661.50 2661.43 375.73 524288 80 5221.98 5222.47 5222.23 382.96 1048576 40 10350.15 10351.85 10351.00 386.40 2097152 20 20246.17 20247.90 20247.04 395.10 4194304 10 40068.15 40076.35 40072.25 399.24 #----------------------------------------------------------------------------- # Benchmarking Exchange # ( #processes = 4 ) # ( 4 additional processes waiting in MPI_Barrier) #----------------------------------------------------------------------------- #bytes #repetitions t_min[usec] t_max[usec] t_avg[usec] Mbytes/sec 0 1000 24.61 24.66 24.64 0.00 1 1000 24.69 24.80 24.74 0.15 2 1000 24.63 24.72 24.69 0.31 4 1000 24.69 24.78 24.72 0.62 8 1000 24.59 24.67 24.65 1.24 16 1000 24.88 25.00 24.95 2.44 32 1000 25.03 25.11 25.07 4.86 64 1000 24.65 24.71 24.68 9.88 128 1000 29.18 29.28 29.24 16.68 256 1000 31.05 31.17 31.12 31.33 512 1000 34.00 34.09 34.05 57.29 1024 1000 42.03 42.09 42.06 92.80 2048 1000 59.82 59.90 59.85 130.43 4096 1000 98.91 98.96 98.95 157.89 8192 1000 174.40 174.49 174.43 179.09 16384 1000 316.27 316.40 316.36 197.54 32768 1000 503.33 503.38 503.36 248.32 65536 640 823.77 824.01 823.88 303.39 131072 320 1461.91 1462.31 1462.14 341.93 262144 160 2784.58 2785.17 2784.88 359.05 524288 80 5429.66 5431.16 5430.32 368.25 1048576 40 10723.11 10725.93 10724.78 372.93 2097152 20 20874.80 20880.13 20876.90 383.14 4194304 10 41175.90 41188.90 41182.44 388.45 #----------------------------------------------------------------------------- # Benchmarking Exchange # ( #processes = 8 ) #----------------------------------------------------------------------------- #bytes #repetitions t_min[usec] t_max[usec] t_avg[usec] Mbytes/sec 0 1000 25.80 25.87 25.83 0.00 1 1000 25.92 26.03 25.97 0.15 2 1000 25.86 25.96 25.93 0.29 4 1000 25.83 25.98 25.90 0.59 8 1000 25.62 25.76 25.69 1.18 16 1000 25.81 25.96 25.89 2.35 32 1000 25.90 26.02 25.97 4.69 64 1000 26.14 26.26 26.20 9.30 128 1000 30.12 30.26 30.17 16.14 256 1000 31.95 32.05 32.01 30.47 512 1000 35.05 35.16 35.10 55.55 1024 1000 42.88 42.98 42.92 90.89 2048 1000 61.25 61.38 61.32 127.28 4096 1000 102.74 102.86 102.80 151.91 8192 1000 178.80 178.90 178.85 174.68 16384 1000 329.09 329.24 329.16 189.83 32768 1000 516.21 516.38 516.30 242.07 65536 640 883.65 883.87 883.74 282.85 131072 320 1610.20 1610.76 1610.46 310.41 262144 160 3214.66 3219.79 3217.33 310.58 524288 80 6066.53 6083.24 6076.51 328.77 1048576 40 11881.79 11886.07 11884.47 336.53 2097152 20 23314.87 23322.25 23317.71 343.02 4194304 10 46258.20 46298.30 46279.22 345.59 #---------------------------------------------------------------- # Benchmarking Allreduce # ( #processes = 2 ) # ( 6 additional processes waiting in MPI_Barrier) #---------------------------------------------------------------- #bytes #repetitions t_min[usec] t_max[usec] t_avg[usec] 0 1000 0.40 0.40 0.40 4 1000 17.55 17.63 17.59 8 1000 17.64 17.73 17.69 16 1000 18.38 18.44 18.41 32 1000 18.73 18.78 18.76 64 1000 19.25 19.25 19.25 128 1000 25.76 25.77 25.77 256 1000 31.22 31.22 31.22 512 1000 37.01 37.05 37.03 1024 1000 49.30 49.32 49.31 2048 1000 73.38 73.43 73.40 4096 1000 109.03 109.08 109.05 8192 1000 182.75 182.86 182.81 16384 1000 311.65 311.66 311.66 32768 1000 526.23 526.32 526.27 65536 640 960.17 960.22 960.20 131072 320 2296.95 2297.01 2296.98 262144 160 5278.33 5278.69 5278.51 524288 80 10961.54 10962.33 10961.93 1048576 40 22345.90 22346.00 22345.95 2097152 20 46120.33 46122.35 46121.34 4194304 10 105679.60 105684.25 105681.93 #---------------------------------------------------------------- # Benchmarking Allreduce # ( #processes = 4 ) # ( 4 additional processes waiting in MPI_Barrier) #---------------------------------------------------------------- #bytes #repetitions t_min[usec] t_max[usec] t_avg[usec] 0 1000 0.40 0.40 0.40 4 1000 34.33 34.34 34.33 8 1000 34.88 34.96 34.93 16 1000 36.23 36.29 36.27 32 1000 36.88 36.91 36.90 64 1000 37.80 37.88 37.86 128 1000 49.74 49.85 49.79 256 1000 61.46 61.55 61.51 512 1000 72.82 72.96 72.87 1024 1000 97.16 97.27 97.20 2048 1000 145.10 145.19 145.15 4096 1000 217.67 217.84 217.73 8192 1000 377.85 378.21 378.04 16384 1000 617.22 617.27 617.24 32768 1000 1044.54 1044.64 1044.58 65536 640 1916.43 1916.51 1916.47 131072 320 4176.40 4176.69 4176.50 262144 160 9850.33 9850.85 9850.57 524288 80 20228.52 20229.37 20228.93 1048576 40 41058.31 41060.01 41059.00 2097152 20 89232.33 89240.40 89235.88 4194304 10 177664.30 177673.35 177667.31 #---------------------------------------------------------------- # Benchmarking Allreduce # ( #processes = 8 ) #---------------------------------------------------------------- #bytes #repetitions t_min[usec] t_max[usec] t_avg[usec] 0 1000 0.40 0.40 0.40 4 1000 51.91 52.10 51.99 8 1000 52.44 52.52 52.47 16 1000 54.77 54.89 54.83 32 1000 55.55 55.66 55.60 64 1000 57.12 57.23 57.18 128 1000 74.21 74.32 74.26 256 1000 93.58 93.71 93.65 512 1000 109.98 110.20 110.10 1024 1000 144.90 145.08 145.00 2048 1000 217.17 217.35 217.27 4096 1000 327.50 327.73 327.61 8192 1000 576.68 577.03 576.89 16384 1000 924.01 924.13 924.05 32768 1000 1563.44 1563.57 1563.51 65536 640 2891.36 2891.57 2891.49 131072 320 6175.67 6175.84 6175.76 262144 160 14118.32 14118.64 14118.51 524288 80 28793.50 28796.21 28794.84 1048576 40 58234.79 58237.01 58235.58 2097152 20 123125.33 123131.65 123129.14 4194304 10 252194.20 252205.20 252199.55 #---------------------------------------------------------------- # Benchmarking Reduce # ( #processes = 2 ) # ( 6 additional processes waiting in MPI_Barrier) #---------------------------------------------------------------- #bytes #repetitions t_min[usec] t_max[usec] t_avg[usec] 0 1000 0.20 0.20 0.20 4 1000 8.91 8.93 8.92 8 1000 9.02 9.04 9.03 16 1000 9.42 9.44 9.43 32 1000 9.52 9.55 9.53 64 1000 9.91 9.98 9.94 128 1000 13.78 13.82 13.80 256 1000 15.99 16.07 16.03 512 1000 18.78 18.82 18.80 1024 1000 24.77 24.82 24.79 2048 1000 36.57 36.64 36.60 4096 1000 54.59 54.62 54.60 8192 1000 90.01 90.13 90.07 16384 1000 194.14 194.21 194.17 32768 1000 338.77 338.89 338.83 65536 640 667.64 667.98 667.81 131072 320 1754.00 1756.05 1755.03 262144 160 4059.67 4073.80 4066.74 524288 80 8561.89 8621.18 8591.53 1048576 40 17158.70 17400.74 17279.72 2097152 20 40478.23 41146.80 40812.51 4194304 10 84090.65 87758.30 85924.47 #---------------------------------------------------------------- # Benchmarking Reduce # ( #processes = 4 ) # ( 4 additional processes waiting in MPI_Barrier) #---------------------------------------------------------------- #bytes #repetitions t_min[usec] t_max[usec] t_avg[usec] 0 1000 0.20 0.20 0.20 4 1000 17.26 17.33 17.30 8 1000 17.69 17.70 17.70 16 1000 18.44 18.50 18.47 32 1000 18.72 18.76 18.74 64 1000 19.26 19.36 19.31 128 1000 26.10 26.17 26.13 256 1000 31.17 31.24 31.20 512 1000 36.59 36.65 36.63 1024 1000 48.35 48.48 48.42 2048 1000 71.74 71.81 71.78 4096 1000 111.31 111.53 111.45 8192 1000 182.59 182.90 182.75 16384 1000 382.17 382.50 382.35 32768 1000 675.02 675.68 675.34 65536 640 1397.20 1399.32 1398.30 131072 320 3176.86 3185.36 3181.28 262144 160 7296.10 7337.93 7317.26 524288 80 15115.58 15289.69 15204.00 1048576 40 30300.82 30995.38 30649.40 2097152 20 66342.22 68695.18 67553.03 4194304 10 131432.45 143453.00 137368.25 #---------------------------------------------------------------- # Benchmarking Reduce # ( #processes = 8 ) #---------------------------------------------------------------- #bytes #repetitions t_min[usec] t_max[usec] t_avg[usec] 0 1000 0.20 0.20 0.20 4 1000 26.45 26.58 26.51 8 1000 26.66 26.79 26.70 16 1000 28.05 28.17 28.11 32 1000 28.22 28.36 28.27 64 1000 29.31 29.41 29.35 128 1000 38.60 38.88 38.71 256 1000 46.32 46.47 46.41 512 1000 54.34 54.52 54.40 1024 1000 71.50 71.76 71.64 2048 1000 106.45 106.69 106.56 4096 1000 167.47 167.81 167.63 8192 1000 275.31 275.86 275.53 16384 1000 571.57 572.45 572.05 32768 1000 1018.44 1019.89 1019.19 65536 640 2022.73 2027.24 2025.08 131072 320 4602.99 4624.20 4613.71 262144 160 10877.24 10983.67 10930.68 524288 80 22487.32 22935.59 22711.45 1048576 40 44814.70 46617.50 45714.75 2097152 20 94286.80 100081.60 97269.47 4194304 10 173543.60 199801.30 186581.84 #---------------------------------------------------------------- # Benchmarking Reduce_scatter # ( #processes = 2 ) # ( 6 additional processes waiting in MPI_Barrier) #---------------------------------------------------------------- #bytes #repetitions t_min[usec] t_max[usec] t_avg[usec] 0 1000 0.30 0.30 0.30 4 1000 19.35 19.36 19.36 8 1000 19.64 19.68 19.66 16 1000 20.09 20.15 20.12 32 1000 20.57 20.66 20.62 64 1000 21.17 21.20 21.19 128 1000 24.96 25.04 25.00 256 1000 30.94 30.98 30.96 512 1000 37.18 37.23 37.21 1024 1000 46.75 46.81 46.78 2048 1000 66.25 66.31 66.28 4096 1000 98.90 98.94 98.92 8192 1000 159.61 159.63 159.62 16384 1000 315.24 315.28 315.26 32768 1000 509.09 509.09 509.09 65536 640 979.32 979.40 979.36 131072 320 2589.96 2590.07 2590.01 262144 160 6157.33 6157.61 6157.47 524288 80 12692.43 12692.52 12692.47 1048576 40 25835.81 25836.96 25836.39 2097152 20 61532.57 61762.48 61647.53 4194304 10 122969.25 123267.20 123118.22 #---------------------------------------------------------------- # Benchmarking Reduce_scatter # ( #processes = 4 ) # ( 4 additional processes waiting in MPI_Barrier) #---------------------------------------------------------------- #bytes #repetitions t_min[usec] t_max[usec] t_avg[usec] 0 1000 0.30 0.30 0.30 4 1000 40.36 40.45 40.41 8 1000 40.73 40.81 40.76 16 1000 41.45 41.48 41.46 32 1000 41.92 41.95 41.94 64 1000 42.77 42.87 42.83 128 1000 49.98 50.08 50.05 256 1000 55.39 55.52 55.46 512 1000 65.42 65.54 65.48 1024 1000 81.24 81.33 81.29 2048 1000 105.97 106.11 106.05 4096 1000 150.14 150.29 150.20 8192 1000 243.58 243.71 243.64 16384 1000 489.88 489.97 489.93 32768 1000 877.13 877.25 877.19 65536 640 1710.80 1711.27 1711.09 131072 320 4080.01 4081.31 4080.84 262144 160 9371.97 9376.30 9374.76 524288 80 19283.17 19297.86 19292.43 1048576 40 38557.89 38619.26 38596.18 2097152 20 85931.30 86448.97 86167.86 4194304 10 175944.15 177407.25 176733.19 #---------------------------------------------------------------- # Benchmarking Reduce_scatter # ( #processes = 8 ) #---------------------------------------------------------------- #bytes #repetitions t_min[usec] t_max[usec] t_avg[usec] 0 1000 0.30 0.30 0.30 4 1000 86.19 86.33 86.25 8 1000 86.62 86.75 86.68 16 1000 87.07 87.16 87.11 32 1000 86.46 86.59 86.53 64 1000 87.19 87.32 87.24 128 1000 97.50 97.67 97.57 256 1000 104.00 104.12 104.07 512 1000 114.23 114.33 114.28 1024 1000 135.72 135.82 135.76 2048 1000 170.36 170.51 170.45 4096 1000 225.75 225.92 225.85 8192 1000 347.84 347.91 347.88 16384 1000 678.75 678.84 678.81 32768 1000 1210.90 1211.08 1210.99 65536 640 2277.37 2277.74 2277.56 131072 320 5489.01 5491.50 5490.26 262144 160 12740.86 12748.73 12745.20 524288 80 25994.99 26021.04 26009.48 1048576 40 52863.55 52959.27 52917.37 2097152 20 112314.42 112984.00 112588.27 4194304 10 231154.85 233254.85 232173.49 #---------------------------------------------------------------- # Benchmarking Allgather # ( #processes = 2 ) # ( 6 additional processes waiting in MPI_Barrier) #---------------------------------------------------------------- #bytes #repetitions t_min[usec] t_max[usec] t_avg[usec] 0 1000 14.71 14.75 14.73 1 1000 14.77 14.78 14.78 2 1000 14.31 14.36 14.34 4 1000 14.52 14.52 14.52 8 1000 13.82 13.91 13.87 16 1000 14.62 14.66 14.64 32 1000 14.50 14.54 14.52 64 1000 15.66 15.67 15.67 128 1000 18.61 18.61 18.61 256 1000 22.44 22.47 22.45 512 1000 24.78 24.81 24.79 1024 1000 31.69 31.72 31.71 2048 1000 46.28 46.34 46.31 4096 1000 76.92 76.92 76.92 8192 1000 134.53 134.54 134.54 16384 1000 189.28 189.31 189.29 32768 1000 317.82 317.82 317.82 65536 640 578.79 578.81 578.80 131072 320 1345.47 1345.67 1345.57 262144 160 3389.69 3389.79 3389.74 524288 80 7156.16 7156.28 7156.22 1048576 40 14677.26 14679.40 14678.33 2097152 20 29553.00 29561.35 29557.18 4194304 10 59442.50 59445.75 59444.12 #---------------------------------------------------------------- # Benchmarking Allgather # ( #processes = 4 ) # ( 4 additional processes waiting in MPI_Barrier) #---------------------------------------------------------------- #bytes #repetitions t_min[usec] t_max[usec] t_avg[usec] 0 1000 41.52 41.60 41.56 1 1000 41.44 41.54 41.49 2 1000 41.92 41.98 41.96 4 1000 41.85 41.99 41.93 8 1000 42.48 42.58 42.54 16 1000 42.92 43.01 42.99 32 1000 43.25 43.35 43.30 64 1000 43.64 43.74 43.67 128 1000 50.91 51.01 50.97 256 1000 59.16 59.33 59.24 512 1000 67.77 67.87 67.81 1024 1000 86.28 86.36 86.30 2048 1000 125.18 125.38 125.26 4096 1000 204.60 204.73 204.66 8192 1000 350.71 350.78 350.74 16384 1000 629.19 629.23 629.21 32768 1000 931.43 931.62 931.54 65536 640 1534.23 1534.54 1534.36 131072 320 3041.43 3041.70 3041.51 262144 160 6589.41 6589.74 6589.59 524288 80 13228.84 13230.86 13229.85 1048576 40 26413.58 26421.04 26417.77 2097152 20 51148.32 51154.07 51151.24 4194304 10 100990.60 101008.00 100998.73 #---------------------------------------------------------------- # Benchmarking Allgather # ( #processes = 8 ) #---------------------------------------------------------------- #bytes #repetitions t_min[usec] t_max[usec] t_avg[usec] 0 1000 102.66 102.72 102.69 1 1000 103.33 103.46 103.39 2 1000 102.97 103.15 103.05 4 1000 103.35 103.41 103.37 8 1000 103.79 103.83 103.81 16 1000 101.89 101.94 101.92 32 1000 104.73 104.79 104.76 64 1000 105.28 105.40 105.37 128 1000 123.09 123.26 123.14 256 1000 140.32 140.45 140.38 512 1000 159.33 159.45 159.40 1024 1000 200.75 200.84 200.80 2048 1000 292.13 292.28 292.22 4096 1000 468.59 468.73 468.65 8192 1000 827.57 827.72 827.65 16384 1000 1293.72 1293.93 1293.82 32768 1000 2014.72 2014.87 2014.79 65536 640 3421.76 3422.03 3421.90 131072 320 6596.39 6597.01 6596.78 262144 160 13223.59 13225.66 13224.73 524288 80 26293.08 26295.14 26294.31 1048576 40 52404.19 52410.67 52409.28 2097152 20 103895.73 104006.90 103955.89 4194304 10 232054.25 232242.95 232169.99 #---------------------------------------------------------------- # Benchmarking Allgatherv # ( #processes = 2 ) # ( 6 additional processes waiting in MPI_Barrier) #---------------------------------------------------------------- #bytes #repetitions t_min[usec] t_max[usec] t_avg[usec] 0 1000 14.27 14.35 14.31 1 1000 14.97 14.98 14.97 2 1000 13.69 13.70 13.70 4 1000 14.21 14.25 14.23 8 1000 14.48 14.60 14.54 16 1000 15.13 15.15 15.14 32 1000 15.03 15.08 15.06 64 1000 15.27 15.38 15.33 128 1000 18.62 18.62 18.62 256 1000 22.42 22.45 22.44 512 1000 24.84 24.85 24.85 1024 1000 31.70 31.72 31.71 2048 1000 46.09 46.18 46.14 4096 1000 77.46 77.53 77.49 8192 1000 131.26 131.29 131.28 16384 1000 189.20 189.26 189.23 32768 1000 316.94 316.98 316.96 65536 640 577.40 577.52 577.46 131072 320 1318.28 1318.29 1318.29 262144 160 3388.08 3388.67 3388.38 524288 80 7152.20 7152.86 7152.53 1048576 40 14674.99 14675.70 14675.34 2097152 20 29538.58 29540.53 29539.55 4194304 10 59376.65 59386.45 59381.55 #---------------------------------------------------------------- # Benchmarking Allgatherv # ( #processes = 4 ) # ( 4 additional processes waiting in MPI_Barrier) #---------------------------------------------------------------- #bytes #repetitions t_min[usec] t_max[usec] t_avg[usec] 0 1000 41.99 42.08 42.05 1 1000 42.32 42.47 42.40 2 1000 42.47 42.58 42.52 4 1000 41.77 41.94 41.88 8 1000 41.77 41.88 41.84 16 1000 43.10 43.14 43.12 32 1000 43.22 43.33 43.26 64 1000 43.40 43.51 43.44 128 1000 50.63 50.76 50.69 256 1000 59.25 59.37 59.30 512 1000 67.91 68.03 67.98 1024 1000 86.14 86.28 86.21 2048 1000 125.33 125.41 125.37 4096 1000 203.95 204.06 204.00 8192 1000 350.96 351.07 351.01 16384 1000 624.53 624.65 624.59 32768 1000 932.58 932.75 932.68 65536 640 1537.46 1537.74 1537.58 131072 320 3042.87 3043.12 3042.96 262144 160 6558.29 6558.94 6558.58 524288 80 13196.78 13198.12 13197.58 1048576 40 26449.59 26456.35 26452.84 2097152 20 51304.80 51316.25 51309.28 4194304 10 100961.70 101017.70 100993.07 #---------------------------------------------------------------- # Benchmarking Allgatherv # ( #processes = 8 ) #---------------------------------------------------------------- #bytes #repetitions t_min[usec] t_max[usec] t_avg[usec] 0 1000 102.54 102.73 102.63 1 1000 102.89 103.03 102.95 2 1000 102.99 103.13 103.05 4 1000 101.30 101.45 101.38 8 1000 101.52 101.57 101.54 16 1000 102.83 102.97 102.92 32 1000 104.11 104.28 104.15 64 1000 104.94 105.03 104.98 128 1000 122.32 122.46 122.40 256 1000 141.39 141.48 141.44 512 1000 158.30 158.46 158.38 1024 1000 199.32 199.45 199.40 2048 1000 290.92 291.02 290.99 4096 1000 468.92 469.01 468.97 8192 1000 820.58 820.70 820.65 16384 1000 1294.29 1294.44 1294.36 32768 1000 2016.07 2016.26 2016.17 65536 640 3419.19 3419.42 3419.33 131072 320 6598.72 6599.36 6599.00 262144 160 13228.06 13229.02 13228.54 524288 80 26305.40 26308.44 26306.59 1048576 40 52354.21 52360.80 52357.81 2097152 20 104052.95 104198.23 104128.40 4194304 10 231862.65 232061.25 231985.97 #---------------------------------------------------------------- # Benchmarking Alltoall # ( #processes = 2 ) # ( 6 additional processes waiting in MPI_Barrier) #---------------------------------------------------------------- #bytes #repetitions t_min[usec] t_max[usec] t_avg[usec] 0 1000 12.77 12.77 12.77 1 1000 13.52 13.57 13.55 2 1000 12.89 12.96 12.93 4 1000 13.61 13.64 13.63 8 1000 12.64 12.65 12.64 16 1000 12.74 12.83 12.79 32 1000 12.95 13.00 12.98 64 1000 13.34 13.36 13.35 128 1000 17.16 17.18 17.17 256 1000 20.66 20.70 20.68 512 1000 23.84 23.84 23.84 1024 1000 29.39 29.49 29.44 2048 1000 42.65 42.71 42.68 4096 1000 71.95 71.99 71.97 8192 1000 121.88 121.91 121.90 16384 1000 185.24 185.33 185.28 32768 1000 315.11 315.12 315.11 65536 640 604.30 604.31 604.31 131072 320 1447.86 1447.87 1447.86 262144 160 3702.07 3702.26 3702.16 524288 80 7454.77 7456.23 7455.50 1048576 40 15004.81 15006.31 15005.56 2097152 20 30123.37 30124.95 30124.16 4194304 10 59627.55 59637.15 59632.35 #---------------------------------------------------------------- # Benchmarking Alltoall # ( #processes = 4 ) # ( 4 additional processes waiting in MPI_Barrier) #---------------------------------------------------------------- #bytes #repetitions t_min[usec] t_max[usec] t_avg[usec] 0 1000 37.03 37.13 37.08 1 1000 37.17 37.27 37.20 2 1000 37.28 37.35 37.31 4 1000 37.22 37.31 37.27 8 1000 37.27 37.38 37.30 16 1000 37.35 37.45 37.40 32 1000 37.02 37.15 37.07 64 1000 37.17 37.23 37.21 128 1000 43.75 43.85 43.79 256 1000 44.45 44.56 44.51 512 1000 48.47 48.56 48.51 1024 1000 56.12 56.15 56.13 2048 1000 76.81 76.87 76.84 4096 1000 149.01 149.15 149.06 8192 1000 277.20 277.32 277.26 16384 1000 653.56 653.79 653.68 32768 1000 1164.54 1164.80 1164.69 65536 640 2251.28 2251.75 2251.59 131072 320 4606.86 4608.13 4607.77 262144 160 9803.09 9806.32 9805.44 524288 80 19163.39 19184.16 19178.05 1048576 40 40158.17 40209.89 40196.11 2097152 20 80128.00 80252.20 80217.86 4194304 10 153502.05 153755.35 153651.21 #---------------------------------------------------------------- # Benchmarking Alltoall # ( #processes = 8 ) #---------------------------------------------------------------- #bytes #repetitions t_min[usec] t_max[usec] t_avg[usec] 0 1000 83.58 83.73 83.66 1 1000 83.83 83.97 83.88 2 1000 83.78 83.92 83.89 4 1000 83.32 83.46 83.37 8 1000 83.14 83.29 83.21 16 1000 82.93 83.00 82.96 32 1000 83.29 83.47 83.38 64 1000 84.21 84.33 84.25 128 1000 95.81 95.94 95.86 256 1000 97.73 97.78 97.76 512 1000 103.87 104.09 103.99 1024 1000 119.10 119.20 119.14 2048 1000 162.78 163.03 162.86 4096 1000 336.29 336.58 336.48 8192 1000 718.12 718.55 718.43 16384 1000 1551.10 1551.54 1551.33 32768 1000 2720.88 2721.58 2721.29 65536 640 5002.29 5003.55 5003.00 131072 320 10442.77 10446.15 10444.41 262144 160 21726.32 21743.51 21736.18 524288 80 44052.15 44123.91 44093.81 1048576 40 91139.78 91403.29 91290.52 2097152 20 184468.40 185534.53 185157.43 4194304 10 405428.90 410998.60 409026.24 #---------------------------------------------------------------- # Benchmarking Bcast # ( #processes = 2 ) # ( 6 additional processes waiting in MPI_Barrier) #---------------------------------------------------------------- #bytes #repetitions t_min[usec] t_max[usec] t_avg[usec] 0 1000 0.20 0.20 0.20 1 1000 8.36 8.45 8.40 2 1000 8.54 8.57 8.55 4 1000 8.54 8.61 8.57 8 1000 8.68 8.72 8.70 16 1000 8.93 8.99 8.96 32 1000 9.15 9.31 9.23 64 1000 9.21 9.21 9.21 128 1000 13.45 13.47 13.46 256 1000 15.20 15.30 15.25 512 1000 18.42 18.47 18.44 1024 1000 25.08 25.19 25.13 2048 1000 37.55 37.57 37.56 4096 1000 57.17 57.29 57.23 8192 1000 96.50 96.57 96.54 16384 1000 117.08 117.08 117.08 32768 1000 187.33 187.43 187.38 65536 640 328.71 328.75 328.73 131072 320 610.27 610.30 610.28 262144 160 1174.04 1174.79 1174.41 524288 80 2304.63 2304.69 2304.66 1048576 40 4569.15 4569.50 4569.33 2097152 20 8919.62 8920.25 8919.94 4194304 10 17591.10 17593.30 17592.20 #---------------------------------------------------------------- # Benchmarking Bcast # ( #processes = 4 ) # ( 4 additional processes waiting in MPI_Barrier) #---------------------------------------------------------------- #bytes #repetitions t_min[usec] t_max[usec] t_avg[usec] 0 1000 0.20 0.20 0.20 1 1000 16.70 16.83 16.75 2 1000 16.72 16.81 16.76 4 1000 16.73 16.85 16.80 8 1000 16.87 16.97 16.94 16 1000 17.31 17.42 17.38 32 1000 17.42 17.48 17.45 64 1000 17.76 17.83 17.80 128 1000 20.75 20.83 20.80 256 1000 20.63 20.72 20.66 512 1000 24.15 24.23 24.17 1024 1000 31.86 31.92 31.89 2048 1000 48.92 49.00 48.97 4096 1000 78.84 79.00 78.92 8192 1000 137.86 138.05 137.97 16384 1000 233.08 233.18 233.11 32768 1000 373.95 374.08 374.01 65536 640 655.78 655.89 655.83 131072 320 1215.27 1215.58 1215.39 262144 160 2342.87 2343.10 2342.97 524288 80 4601.94 4602.66 4602.32 1048576 40 9123.56 9125.14 9124.29 2097152 20 17807.18 17811.70 17810.08 4194304 10 35162.75 35177.15 35167.46 #---------------------------------------------------------------- # Benchmarking Bcast # ( #processes = 8 ) #---------------------------------------------------------------- #bytes #repetitions t_min[usec] t_max[usec] t_avg[usec] 0 1000 0.20 0.20 0.20 1 1000 21.68 21.78 21.73 2 1000 21.72 21.80 21.75 4 1000 21.60 21.75 21.70 8 1000 21.73 21.90 21.82 16 1000 22.13 22.21 22.18 32 1000 22.19 22.40 22.31 64 1000 22.53 22.64 22.60 128 1000 26.38 26.46 26.42 256 1000 30.22 30.41 30.33 512 1000 31.80 31.87 31.83 1024 1000 38.73 38.93 38.86 2048 1000 57.44 57.60 57.53 4096 1000 97.74 97.99 97.89 8192 1000 177.96 178.23 178.10 16384 1000 349.55 349.64 349.58 32768 1000 561.20 561.31 561.26 65536 640 985.23 985.36 985.29 131072 320 1827.70 1828.10 1827.91 262144 160 3528.35 3529.23 3528.69 524288 80 6938.29 6940.14 6939.43 1048576 40 13762.18 13767.94 13764.25 2097152 20 26797.78 26803.77 26799.81 4194304 10 52796.35 52819.20 52808.61 #--------------------------------------------------- # Benchmarking Barrier # ( #processes = 2 ) # ( 6 additional processes waiting in MPI_Barrier) #--------------------------------------------------- #repetitions t_min[usec] t_max[usec] t_avg[usec] 1000 13.15 13.18 13.17 #--------------------------------------------------- # Benchmarking Barrier # ( #processes = 4 ) # ( 4 additional processes waiting in MPI_Barrier) #--------------------------------------------------- #repetitions t_min[usec] t_max[usec] t_avg[usec] 1000 28.49 28.58 28.53 #--------------------------------------------------- # Benchmarking Barrier # ( #processes = 8 ) #--------------------------------------------------- #repetitions t_min[usec] t_max[usec] t_avg[usec] 1000 45.04 45.19 45.12 #===================================================== # # Thanks for using PMB2.2 # # The Pallas team kindly requests that you # give us as much feedback for PMB as possible. # # It would be very helpful when you sent the # output tables of your run(s) of PMB to # # ####################### # # # # # pmb@pallas.com # # # # # ####################### # # You might also add # # - personal information (institution, motivation # for using PMB) # - basic information about the machine you used # (number of CPUs, processor type e.t.c.) # #=====================================================