25 June 2010 12:00:47 PM FFT_OPEN_MP C++/OpenMP version Demonstrate an implementation of the Fast Fourier Transform of a complex data vector, using OpenMP for parallel execution. Number of processors available = 8 Number of threads = 1 Accuracy check: FFT ( FFT ( X(1:N) ) ) == N * X(1:N) N NITS Error Time Time/Call MFLOPS 2 10000 7.85908e-17 1 4.83894e-07 20.6657 4 10000 1.20984e-16 1 9.59241e-07 41.6996 8 10000 6.8208e-17 1 1.1819e-06 101.531 16 10000 1.43867e-16 1 1.46581e-06 218.309 32 1000 1.33121e-16 1 2.25353e-06 354.998 64 1000 1.77654e-16 1 3.57497e-06 537.068 128 1000 1.92904e-16 1 6.14846e-06 728.638 256 1000 2.09232e-16 1 1.2354e-05 828.88 512 100 1.92749e-16 1 2.43294e-05 947.001 1024 100 2.30861e-16 1 5.51391e-05 928.561 2048 100 2.44451e-16 1 0.00011568 973.725 4096 100 2.48327e-16 1 0.000257225 955.428 8192 10 2.57505e-16 1 0.000543892 979.017 16384 10 2.72935e-16 1 0.00120566 951.247 32768 10 2.91185e-16 1 0.00250649 980.493 65536 10 2.82726e-16 1 0.005505 952.385 131072 1 3.14826e-16 1 0.011389 978.234 262144 1 3.222e-16 1 0.0265185 889.68 524288 1 3.27877e-16 1 0.0568734 875.758 1048576 1 3.28641e-16 1 0.124539 841.969 2097152 1 3.50863e-16 1 0.252566 871.854 4194304 1 3.56312e-16 1 0.561061 822.324 8388608 1 3.71083e-16 1 1.10563 872.526 16777216 1 3.65694e-16 1 2.34929 856.969 33554432 1 3.81727e-16 1 4.82616 869.077 FFT_OPEN_MP: Normal end of execution. 25 June 2010 12:01:31 PM 25 June 2010 12:01:31 PM FFT_OPEN_MP C++/OpenMP version Demonstrate an implementation of the Fast Fourier Transform of a complex data vector, using OpenMP for parallel execution. Number of processors available = 8 Number of threads = 2 Accuracy check: FFT ( FFT ( X(1:N) ) ) == N * X(1:N) N NITS Error Time Time/Call MFLOPS 2 10000 7.85908e-17 1 1.13806e-06 8.78692 4 10000 1.20984e-16 1 1.8487e-06 21.6369 8 10000 6.8208e-17 1 2.77435e-06 43.2534 16 10000 1.43867e-16 1 3.82659e-06 83.6253 32 1000 1.33121e-16 1 4.95195e-06 161.552 64 1000 1.77654e-16 1 6.65808e-06 288.372 128 1000 1.92904e-16 1 8.79002e-06 509.669 256 1000 2.09232e-16 1 1.34109e-05 763.557 512 100 1.92749e-16 1 2.15948e-05 1066.93 1024 100 2.30861e-16 1 3.95799e-05 1293.59 2048 100 2.44451e-16 1 7.26104e-05 1551.29 4096 100 2.48327e-16 1 0.000155416 1581.31 8192 10 2.57505e-16 1 0.000305951 1740.41 16384 10 2.72935e-16 1 0.000684297 1676 32768 10 2.91185e-16 1 0.00136725 1797.48 65536 10 2.82726e-16 1 0.0030316 1729.41 131072 1 3.14826e-16 1 0.00603497 1846.09 262144 1 3.222e-16 1 0.0154309 1528.94 524288 1 3.27877e-16 1 0.0345345 1442.25 1048576 1 3.28641e-16 1 0.0849481 1234.37 2097152 1 3.50863e-16 1 0.175235 1256.6 4194304 1 3.56312e-16 1 0.369006 1250.31 8388608 1 3.71083e-16 1 0.689638 1398.84 16777216 1 3.65694e-16 1 1.60676 1252.99 33554432 1 3.81727e-16 1 3.14231 1334.78 FFT_OPEN_MP: Normal end of execution. 25 June 2010 12:02:02 PM 25 June 2010 12:02:02 PM FFT_OPEN_MP C++/OpenMP version Demonstrate an implementation of the Fast Fourier Transform of a complex data vector, using OpenMP for parallel execution. Number of processors available = 8 Number of threads = 4 Accuracy check: FFT ( FFT ( X(1:N) ) ) == N * X(1:N) N NITS Error Time Time/Call MFLOPS 2 10000 7.85908e-17 1 4.23555e-06 2.36097 4 10000 1.20984e-16 1 2.16335e-06 18.4898 8 10000 6.8208e-17 1 3.49215e-06 34.3628 16 10000 1.43867e-16 1 4.49915e-06 71.1246 32 1000 1.33121e-16 1 5.96297e-06 134.161 64 1000 1.77654e-16 1 8.08001e-06 237.624 128 1000 1.92904e-16 1 1.0053e-05 445.636 256 1000 2.09232e-16 1 1.4522e-05 705.139 512 100 1.92749e-16 1 2.04849e-05 1124.73 1024 100 2.30861e-16 1 3.48353e-05 1469.77 2048 100 2.44451e-16 1 5.61655e-05 2005.5 4096 100 2.48327e-16 1 0.00011642 2110.98 8192 10 2.57505e-16 1 0.00021255 2505.2 16384 10 2.72935e-16 1 0.000469947 2440.45 32768 10 2.91185e-16 1 0.000897896 2737.06 65536 10 2.82726e-16 1 0.00202751 2585.87 131072 1 3.14826e-16 1 0.00389457 2860.68 262144 1 3.222e-16 1 0.0116426 2026.44 524288 1 3.27877e-16 1 0.0295775 1683.96 1048576 1 3.28641e-16 1 0.0701466 1494.84 2097152 1 3.50863e-16 1 0.133421 1650.42 4194304 1 3.56312e-16 1 0.325946 1415.49 8388608 1 3.71083e-16 1 0.56374 1711.23 16777216 1 3.65694e-16 1 1.25261 1607.25 33554432 1 3.81727e-16 1 2.4218 1731.9 FFT_OPEN_MP: Normal end of execution. 25 June 2010 12:02:28 PM 25 June 2010 12:02:28 PM FFT_OPEN_MP C++/OpenMP version Demonstrate an implementation of the Fast Fourier Transform of a complex data vector, using OpenMP for parallel execution. Number of processors available = 8 Number of threads = 8 Accuracy check: FFT ( FFT ( X(1:N) ) ) == N * X(1:N) N NITS Error Time Time/Call MFLOPS 2 10000 7.85908e-17 1 2.89975e-06 3.44857 4 10000 1.20984e-16 1 5.60236e-06 7.13985 8 10000 6.8208e-17 1 8.57916e-06 13.9874 16 10000 1.43867e-16 1 1.17243e-05 27.2936 32 1000 1.33121e-16 1 1.50925e-05 53.0065 64 1000 1.77654e-16 1 1.84366e-05 104.141 128 1000 1.92904e-16 1 2.19555e-05 204.049 256 1000 2.09232e-16 1 2.93715e-05 348.637 512 100 1.92749e-16 1 3.76558e-05 611.857 1024 100 2.30861e-16 1 5.50807e-05 929.546 2048 100 2.44451e-16 1 7.635e-05 1475.31 4096 100 2.48327e-16 1 0.000133801 1836.76 8192 10 2.57505e-16 1 0.000229204 2323.17 16384 10 2.72935e-16 1 0.00048331 2372.97 32768 10 2.91185e-16 1 0.000885451 2775.54 65536 10 2.82726e-16 1 0.00196925 2662.37 131072 1 3.14826e-16 1 0.00368595 3022.59 262144 1 3.222e-16 1 0.00911701 2587.8 524288 1 3.27877e-16 1 0.0235045 2119.06 1048576 1 3.28641e-16 1 0.057775 1814.93 2097152 1 3.50863e-16 1 0.113741 1935.99 4194304 1 3.56312e-16 1 0.254713 1811.35 8388608 1 3.71083e-16 1 0.484306 1991.9 16777216 1 3.65694e-16 1 1.07215 1877.78 33554432 1 3.81727e-16 1 2.05707 2038.97 FFT_OPEN_MP: Normal end of execution. 25 June 2010 12:02:51 PM