SencRecv Performance Comparison

SencRecv is one of PMB's Parallel Transfer Benchmarks. Based on MPI_Sendrecv, the processes form a periodic communication
chain. Each process sends to the right and receives from the left neighbor in the chain.

The turnover count is 2 messages per sample (1 in, 1 out) for each process.

Sendrecv is equivalent with the Cshift benchmark and, in case of 2 processes, the PingPing benchmark of PMB1.x. For 2 processes, it will report
the bi-directional bandwidth of the system, as obtained by the (optimized) MPI_Sendrecv function.

2-node Cluster Results
4-node Cluster Results
8-node Cluster Results
กก

2-node Latency (millisecond) Performance Results

#bytes  #reps   linux           dom0            domU(noSMP)     domU(SMP)
--------------------------------------------------------------------------

2-node Throughput (MB/s) Performance Results

#bytes  #reps   linux           dom0            domU(noSMP)     domU(SMP)
--------------------------------------------------------------------------

4-node Latency (millisecond) Performance Results

#bytes  #reps   linux           dom0            domU(noSMP)     domU(SMP)
--------------------------------------------------------------------------
กก

4-node Throughput (MB/s) Performance Results

#bytes  #reps   linux           dom0            domU(noSMP)     domU(SMP)
--------------------------------------------------------------------------

8-node Latency (millisecond) Performance Results


#bytes  #reps   linux           dom0            domU(SMP)     domU(noSMP)
--------------------------------------------------------------------------
0       1000       1977.92         1933.71         2628.84         3329.24
1       1000       1606.17          859.68         2954.64         3329.35
2       1000       1687.79          724.26         2764.47         3708.08
4       1000       1613.15          426.52         2481.06         3540.98
8       1000        384.06          282.39         2385.76         3326.26
16      1000        178.31          280.87         2739.28         3104.53
32      1000        183.47          284.91         2270.88         2835.78
64      1000        191.11          294.97         1992.58         3310.37
128     1000        203.41          308.48          588.25         3445.02
256     1000        226.73          336.54          434.53         1391.99
512     1000        288.57          405.45          491.32         1111.03
1024    1000        375.39          516.08          586.76         1260.99
2048    1000        498.92          739.06          764.88         2007.58
4096    1000        722.78          925.39          984.56         2457.22
8192    1000       1172.61         1411.11         2238.01         4172.41
16384   1000       2089.53         2313.20         5906.05         9971.30
32768   1000       6284.80         5554.05        10792.55        24828.09
65536   640       24722.79        13471.37        30270.34        64670.69
131072  320       32415.03        32664.81        32942.32       147792.17
262144  160       65917.24        61949.30        66632.82       207830.00
524288  80       123976.45       129890.96       131876.07       259392.54
1048576 40       261248.37       263271.88       245035.49       389769.31
2097152 20       500725.41       525550.69       565333.27       883586.22
4194304 10      1021610.90      1038750.95      1087948.76      1851702.73

8-node Throughput (MB/sec) Performance Results


#bytes  #reps   linux           dom0            domU(SMP)     domU(noSMP)
--------------------------------------------------------------------------
0       1000          0.00            0.00            0.00            0.00
1       1000          0.00            0.00            0.00            0.00
2       1000          0.00            0.01            0.00            0.00
4       1000          0.00            0.02            0.00            0.00
8       1000          0.04            0.05            0.01            0.00
16      1000          0.17            0.11            0.01            0.01
32      1000          0.33            0.21            0.03            0.02
64      1000          0.64            0.41            0.07            0.04
128     1000          1.20            0.79            0.49            0.07
256     1000          2.15            1.45            1.12            0.36
512     1000          3.38            2.41            1.99            0.88
1024    1000          5.20            3.78            3.33            1.56
2048    1000          7.83            5.28            5.10            1.97
4096    1000         10.80            8.44            7.93            3.36
8192    1000         13.33           11.07            7.06            3.63
16384   1000         14.95           13.51            5.32            3.14
32768   1000          9.94           11.25            5.78            2.52
65536   640           5.05            9.28            4.13            1.92
131072  320           7.70            7.64            7.58            1.69
262144  160           7.56            8.05            7.47            2.39
524288  80            8.08            7.64            7.54            3.82
1048576 40            7.55            7.49            8.09            5.00
2097152 20            7.80            7.39            6.85            4.39
4194304 10            7.40            7.29            6.98            3.99