
================================================================
  Antenna Interpolation Benchmark
  Typical calling pattern: n_ang=1, per-element rotation
  Comparing scalar vs AVX2 implementation
================================================================

========================================================================
  S1: float  n_out=1   n_ang=1  grid=5deg (73x37)
  Pattern: 73 x 37 x 1   n_out: 1   n_ang: 1
  per_element_rotation: YES   dtype: float   n_warmup: 500000   n_reps: 1000
========================================================================

  PERFORMANCE (per call):
    Original :      0.0038 ms   (    0.26 Msamples/s)
    AVX2     :      0.0009 ms   (    1.11 Msamples/s)
    Speedup  :        4.26x

  ACCURACY (AVX2 vs scalar reference):
    Output         |    avg |err|    max |err| |  avg ULP    max ULP |      avg rel      max rel
    --------------------------------------------------------------------------------------------------
    V_re           |    1.788e-07    1.788e-07 |      3.0          3 |    1.966e-07    1.966e-07
    V_im           |    1.192e-07    1.192e-07 |      4.0          4 |    3.288e-07    3.288e-07
    H_re           |    5.960e-08    5.960e-08 |      4.0          4 |    3.653e-07    3.653e-07
    H_im           |    2.235e-08    2.235e-08 |      3.0          3 |    1.936e-07    1.936e-07
    dist           |    0.000e+00    0.000e+00 |      0.0          0 |    0.000e+00    0.000e+00
    az_loc         |    1.490e-08    1.490e-08 |      1.0          1 |    7.024e-08    7.024e-08
    el_loc         |    0.000e+00    0.000e+00 |      0.0          0 |    0.000e+00    0.000e+00
    gamma          |    2.980e-08    2.980e-08 |      4.0          4 |    3.355e-07    3.355e-07

  COMPLEX FIELD ACCURACY:
    field      avg amp err  max amp err   avg ph err   max ph err  avg ULP(re)  max ULP(re)
    ----------------------------------------------------------------------------------------
    V (theta)     2.102e-07    2.102e-07    4.548e-08    4.548e-08          3.0            3
    H (phi)       6.157e-08    6.157e-08    8.096e-08    8.096e-08          4.0            4

========================================================================
  S2: float  n_out=8   n_ang=1  grid=5deg (73x37)
  Pattern: 73 x 37 x 8   n_out: 8   n_ang: 1
  per_element_rotation: YES   dtype: float   n_warmup: 100000   n_reps: 1000
========================================================================

  PERFORMANCE (per call):
    Original :      0.0074 ms   (    1.09 Msamples/s)
    AVX2     :      0.0018 ms   (    4.38 Msamples/s)
    Speedup  :        4.03x

  ACCURACY (AVX2 vs scalar reference):
    Output         |    avg |err|    max |err| |  avg ULP    max ULP |      avg rel      max rel
    --------------------------------------------------------------------------------------------------
    V_re           |    9.779e-08    3.874e-07 |      3.1         13 |    2.208e-07    8.037e-07
    V_im           |    1.267e-07    4.172e-07 |      2.4          7 |    1.929e-07    5.765e-07
    H_re           |    4.098e-08    1.490e-07 |      3.8         12 |    3.624e-07    1.252e-06
    H_im           |    4.377e-08    1.490e-07 |      2.1          5 |    1.577e-07    3.856e-07
    dist           |    7.451e-08    4.768e-07 |      0.5          2 |    3.517e-08    1.599e-07
    az_loc         |    1.304e-08    2.235e-08 |      1.4          3 |    1.119e-07    3.042e-07
    el_loc         |    4.773e-09    1.118e-08 |      1.5          4 |    1.436e-07    4.340e-07
    gamma          |    2.421e-08    4.470e-08 |      1.5          4 |    1.171e-07    3.355e-07

  COMPLEX FIELD ACCURACY:
    field      avg amp err  max amp err   avg ph err   max ph err  avg ULP(re)  max ULP(re)
    ----------------------------------------------------------------------------------------
    V (theta)     1.617e-07    5.620e-07    3.677e-08    1.048e-07          3.1           13
    H (phi)       5.850e-08    2.097e-07    5.269e-08    1.235e-07          3.8           12

========================================================================
  S3: float  n_out=64  n_ang=1  grid=5deg (73x37)
  Pattern: 73 x 37 x 64   n_out: 64   n_ang: 1
  per_element_rotation: YES   dtype: float   n_warmup: 20000   n_reps: 500
========================================================================

  PERFORMANCE (per call):
    Original :      0.0277 ms   (    2.31 Msamples/s)
    AVX2     :      0.0052 ms   (   12.39 Msamples/s)
    Speedup  :        5.37x

  ACCURACY (AVX2 vs scalar reference):
    Output         |    avg |err|    max |err| |  avg ULP    max ULP |      avg rel      max rel
    --------------------------------------------------------------------------------------------------
    V_re           |    6.883e-08    3.874e-07 |     25.2        673 |    1.962e-06    4.236e-05
    V_im           |    7.178e-08    4.172e-07 |     13.3        301 |    9.733e-07    1.833e-05
    H_re           |    1.012e-07    3.576e-07 |      7.2        182 |    5.988e-07    1.589e-05
    H_im           |    8.443e-08    3.576e-07 |      3.9         31 |    3.150e-07    2.646e-06
    dist           |    4.787e-07    1.907e-06 |      0.6          2 |    5.087e-08    2.126e-07
    az_loc         |    3.177e-08    1.192e-07 |      1.5         15 |    1.190e-07    9.028e-07
    el_loc         |    3.220e-08    2.384e-07 |      1.1         14 |    1.017e-07    1.242e-06
    gamma          |    5.332e-08    2.384e-07 |      0.5          4 |    4.355e-08    3.355e-07

  COMPLEX FIELD ACCURACY:
    field      avg amp err  max amp err   avg ph err   max ph err  avg ULP(re)  max ULP(re)
    ----------------------------------------------------------------------------------------
    V (theta)     9.128e-08    5.620e-07    1.964e-07    1.002e-06         25.2          673
    H (phi)       1.209e-07    3.797e-07    9.412e-08    8.618e-07          7.2          182

========================================================================
  S4: float  n_out=8   n_ang=1  grid=1deg (361x181)
  Pattern: 361 x 181 x 8   n_out: 8   n_ang: 1
  per_element_rotation: YES   dtype: float   n_warmup: 50000   n_reps: 1000
========================================================================

  PERFORMANCE (per call):
    Original :      0.3527 ms   (    0.02 Msamples/s)
    AVX2     :      0.0015 ms   (    5.43 Msamples/s)
    Speedup  :      239.55x

  ACCURACY (AVX2 vs scalar reference):
    Output         |    avg |err|    max |err| |  avg ULP    max ULP |      avg rel      max rel
    --------------------------------------------------------------------------------------------------
    V_re           |    9.593e-08    1.490e-07 |      3.6         11 |    3.164e-07    9.572e-07
    V_im           |    6.706e-08    1.192e-07 |      1.1          2 |    8.904e-08    2.045e-07
    H_re           |    3.166e-08    7.451e-08 |      3.1          5 |    2.809e-07    4.672e-07
    H_im           |    3.446e-08    8.941e-08 |      1.9          4 |    1.378e-07    3.394e-07
    dist           |    7.451e-08    4.768e-07 |      0.5          2 |    3.517e-08    1.599e-07
    az_loc         |    1.304e-08    2.235e-08 |      1.4          3 |    1.119e-07    3.042e-07
    el_loc         |    4.773e-09    1.118e-08 |      1.5          4 |    1.436e-07    4.340e-07
    gamma          |    2.421e-08    4.470e-08 |      1.5          4 |    1.171e-07    3.355e-07

  COMPLEX FIELD ACCURACY:
    field      avg amp err  max amp err   avg ph err   max ph err  avg ULP(re)  max ULP(re)
    ----------------------------------------------------------------------------------------
    V (theta)     6.101e-08    1.669e-07    9.119e-08    1.792e-07          3.6           11
    H (phi)       4.293e-08    7.411e-08    6.638e-08    1.198e-07          3.1            5

========================================================================
  S5: float  n_out=64  n_ang=1  grid=1deg (361x181)
  Pattern: 361 x 181 x 64   n_out: 64   n_ang: 1
  per_element_rotation: YES   dtype: float   n_warmup: 10000   n_reps: 500
========================================================================

  PERFORMANCE (per call):
    Original :      0.0475 ms   (    1.35 Msamples/s)
    AVX2     :      0.0058 ms   (   11.05 Msamples/s)
    Speedup  :        8.19x

  ACCURACY (AVX2 vs scalar reference):
    Output         |    avg |err|    max |err| |  avg ULP    max ULP |      avg rel      max rel
    --------------------------------------------------------------------------------------------------
    V_re           |    6.352e-08    2.384e-07 |     24.7        629 |    2.229e-06    4.709e-05
    V_im           |    5.186e-08    2.384e-07 |      5.0         40 |    4.217e-07    3.156e-06
    H_re           |    8.174e-08    2.980e-07 |      5.9        120 |    4.790e-07    1.048e-05
    H_im           |    6.598e-08    2.980e-07 |      2.7         21 |    2.130e-07    1.827e-06
    dist           |    4.787e-07    1.907e-06 |      0.6          2 |    5.087e-08    2.126e-07
    az_loc         |    3.177e-08    1.192e-07 |      1.5         15 |    1.190e-07    9.028e-07
    el_loc         |    3.220e-08    2.384e-07 |      1.1         14 |    1.017e-07    1.242e-06
    gamma          |    5.332e-08    2.384e-07 |      0.5          4 |    4.355e-08    3.355e-07

  COMPLEX FIELD ACCURACY:
    field      avg amp err  max amp err   avg ph err   max ph err  avg ULP(re)  max ULP(re)
    ----------------------------------------------------------------------------------------
    V (theta)     6.980e-08    2.380e-07    1.530e-07    7.168e-07         24.7          629
    H (phi)       9.195e-08    3.405e-07    8.888e-08    5.951e-07          5.9          120

========================================================================
  S6: double n_out=8   n_ang=1  grid=5deg (73x37)
  Pattern: 73 x 37 x 8   n_out: 8   n_ang: 1
  per_element_rotation: YES   dtype: double   n_warmup: 100000   n_reps: 1000
========================================================================

  PERFORMANCE (per call):
    Original :      0.0077 ms   (    1.04 Msamples/s)
    AVX2     :      0.0011 ms   (    6.96 Msamples/s)
    Speedup  :        6.71x

  ACCURACY (AVX2 vs scalar reference):
    Output         |    avg |err|    max |err| |  avg ULP    max ULP |      avg rel      max rel
    --------------------------------------------------------------------------------------------------
    V_re           |    9.813e-08    2.661e-07 |      2.9          6 |    2.201e-07    3.693e-07
    V_im           |    1.896e-07    3.774e-07 |      3.2          6 |    2.572e-07    5.215e-07
    H_re           |    1.826e-08    5.288e-08 |      1.8          5 |    1.595e-07    4.766e-07
    H_im           |    3.740e-08    1.039e-07 |      1.6          3 |    1.278e-07    2.690e-07
    dist           |    4.369e-07    1.010e-06 |      3.0          4 |    2.125e-07    3.395e-07
    az_loc         |    1.392e-08    2.844e-08 |      1.5          2 |    9.950e-08    1.662e-07
    el_loc         |    2.640e-08    3.727e-08 |     10.2         31 |    1.018e-06    3.325e-06
    gamma          |    3.251e-08    6.795e-08 |      1.5          4 |    1.280e-07    2.589e-07

  COMPLEX FIELD ACCURACY:
    field      avg amp err  max amp err   avg ph err   max ph err  avg ULP(re)  max ULP(re)
    ----------------------------------------------------------------------------------------
    V (theta)     2.228e-07    4.128e-07    2.132e-08    7.018e-08          2.9            6
    H (phi)       4.273e-08    1.032e-07    2.830e-08    7.621e-08          1.8            5

========================================================================
  S7: float  n_out=1   n_ang=500  grid=5deg (73x37)
  Pattern: 73 x 37 x 1   n_out: 1   n_ang: 500
  per_element_rotation: YES   dtype: float   n_warmup: 1000   n_reps: 500
========================================================================

  PERFORMANCE (per call):
    Original :      0.0115 ms   (   43.33 Msamples/s)
    AVX2     :      0.1465 ms   (    3.41 Msamples/s)
    Speedup  :        0.08x

  ACCURACY (AVX2 vs scalar reference):
    Output         |    avg |err|    max |err| |  avg ULP    max ULP |      avg rel      max rel
    --------------------------------------------------------------------------------------------------
    V_re           |    1.009e-07    4.768e-07 |      2.0          9 |    1.622e-07    6.821e-07
    V_im           |    1.194e-07    4.768e-07 |      2.2          9 |    1.741e-07    7.263e-07
    H_re           |    2.465e-08    1.043e-07 |      1.7          7 |    1.512e-07    6.518e-07
    H_im           |    4.092e-08    1.490e-07 |      2.9         10 |    2.437e-07    8.317e-07
    dist           |    0.000e+00    0.000e+00 |      0.0          0 |    0.000e+00    0.000e+00
    az_loc         |    2.971e-08    1.192e-07 |      0.7          4 |    6.301e-08    2.586e-07
    el_loc         |    6.288e-09    5.960e-08 |      0.3          2 |    2.733e-08    2.212e-07
    gamma          |    1.333e-08    6.333e-08 |      2.6         17 |    2.079e-07    1.291e-06

  COMPLEX FIELD ACCURACY:
    field      avg amp err  max amp err   avg ph err   max ph err  avg ULP(re)  max ULP(re)
    ----------------------------------------------------------------------------------------
    V (theta)     1.440e-07    5.490e-07    7.113e-08    3.332e-07          2.0            9
    H (phi)       4.420e-08    1.603e-07    7.813e-08    2.873e-07          1.7            7

========================================================================
  S8: float  n_out=1   n_ang=2000 grid=5deg (73x37)
  Pattern: 73 x 37 x 1   n_out: 1   n_ang: 2000
  per_element_rotation: YES   dtype: float   n_warmup: 1000   n_reps: 500
========================================================================

  PERFORMANCE (per call):
    Original :      0.0336 ms   (   59.56 Msamples/s)
    AVX2     :      0.0855 ms   (   23.38 Msamples/s)
    Speedup  :        0.39x

  ACCURACY (AVX2 vs scalar reference):
    Output         |    avg |err|    max |err| |  avg ULP    max ULP |      avg rel      max rel
    --------------------------------------------------------------------------------------------------
    V_re           |    9.038e-08    4.768e-07 |      6.4       3365 |    5.248e-07    2.402e-04
    V_im           |    1.203e-07    5.960e-07 |     20.5       8867 |    1.768e-06    8.337e-04
    H_re           |    5.572e-08    3.874e-07 |      2.6         26 |    2.271e-07    1.758e-06
    H_im           |    8.757e-08    4.172e-07 |      3.2         14 |    2.681e-07    1.091e-06
    dist           |    0.000e+00    0.000e+00 |      0.0          0 |    0.000e+00    0.000e+00
    az_loc         |    4.838e-08    2.384e-07 |      0.6          4 |    4.502e-08    2.586e-07
    el_loc         |    3.440e-08    2.384e-07 |      0.6          3 |    5.070e-08    3.353e-07
    gamma          |    2.603e-08    2.831e-07 |     20.5       5952 |    1.664e-06    4.750e-04

  COMPLEX FIELD ACCURACY:
    field      avg amp err  max amp err   avg ph err   max ph err  avg ULP(re)  max ULP(re)
    ----------------------------------------------------------------------------------------
    V (theta)     1.240e-07    5.490e-07    1.113e-07    5.864e-07          6.4         3365
    H (phi)       7.692e-08    3.686e-07    1.493e-07    6.769e-07          2.6           26

========================================================================
  S9: float  n_out=64  n_ang=500  grid=5deg (73x37)
  Pattern: 73 x 37 x 64   n_out: 64   n_ang: 500
  per_element_rotation: YES   dtype: float   n_warmup: 2000   n_reps: 200
========================================================================

  PERFORMANCE (per call):
    Original :      0.4344 ms   (   73.66 Msamples/s)
    AVX2     :      0.0775 ms   (  413.12 Msamples/s)
    Speedup  :        5.61x

  ACCURACY (AVX2 vs scalar reference):
    Output         |    avg |err|    max |err| |  avg ULP    max ULP |      avg rel      max rel
    --------------------------------------------------------------------------------------------------
    V_re           |    6.810e-08    1.311e-06 |     11.8      21120 |    9.875e-07    2.020e-03
    V_im           |    7.312e-08    2.220e-06 |     15.4      93991 |    1.256e-06    6.058e-03
    H_re           |    8.217e-08    8.577e-07 |      5.8       5918 |    4.756e-07    4.783e-04
    H_im           |    8.569e-08    8.941e-07 |      7.4      13544 |    6.105e-07    1.477e-03
    dist           |    5.382e-07    5.722e-06 |      0.7          5 |    5.146e-08    3.548e-07
    az_loc         |    2.940e-08    2.384e-07 |      2.6      12841 |    2.132e-07    7.713e-04
    el_loc         |    3.273e-08    3.576e-07 |     13.1     314834 |    1.001e-06    2.414e-02
    gamma          |    5.638e-08    4.768e-07 |      0.6         17 |    4.896e-08    1.291e-06

  COMPLEX FIELD ACCURACY:
    field      avg amp err  max amp err   avg ph err   max ph err  avg ULP(re)  max ULP(re)
    ----------------------------------------------------------------------------------------
    V (theta)     9.421e-08    7.784e-07    1.379e-07    6.448e-06         11.8        21120
    H (phi)       1.103e-07    8.829e-07    8.544e-08    1.183e-06          5.8         5918

========================================================================
  S10: float  n_out=64  n_ang=2000 grid=5deg (73x37)
  Pattern: 73 x 37 x 64   n_out: 64   n_ang: 2000
  per_element_rotation: YES   dtype: float   n_warmup: 1000   n_reps: 100
========================================================================

  PERFORMANCE (per call):
    Original :      1.7585 ms   (   72.79 Msamples/s)
    AVX2     :      0.2693 ms   (  475.26 Msamples/s)
    Speedup  :        6.53x

  ACCURACY (AVX2 vs scalar reference):
    Output         |    avg |err|    max |err| |  avg ULP    max ULP |      avg rel      max rel
    --------------------------------------------------------------------------------------------------
    V_re           |    7.317e-08    1.311e-06 |     10.3      44643 |    8.658e-07    3.703e-03
    V_im           |    7.423e-08    2.220e-06 |     10.4      93991 |    8.506e-07    6.058e-03
    H_re           |    7.670e-08    8.577e-07 |     21.2     479840 |    1.729e-06    4.068e-02
    H_im           |    7.750e-08    8.941e-07 |      6.4      31234 |    5.272e-07    2.450e-03
    dist           |    5.693e-07    5.722e-06 |      0.8       2478 |    6.867e-08    1.833e-04
    az_loc         |    3.088e-08    2.384e-07 |      1.7      12841 |    1.403e-07    7.713e-04
    el_loc         |    2.767e-08    3.576e-07 |      4.8     314834 |    3.793e-07    2.414e-02
    gamma          |    4.833e-08    4.768e-07 |      2.4      15737 |    1.992e-07    1.629e-03

  COMPLEX FIELD ACCURACY:
    field      avg amp err  max amp err   avg ph err   max ph err  avg ULP(re)  max ULP(re)
    ----------------------------------------------------------------------------------------
    V (theta)     9.548e-08    7.784e-07    1.097e-07    6.448e-06         10.3        44643
    H (phi)       9.865e-08    8.829e-07    9.459e-08    1.261e-06         21.2       479840

========================================================================
  S11: float  n_out=1   n_ang=500  grid=1deg (361x181)
  Pattern: 361 x 181 x 1   n_out: 1   n_ang: 500
  per_element_rotation: YES   dtype: float   n_warmup: 500   n_reps: 200
========================================================================

  PERFORMANCE (per call):
    Original :      0.0159 ms   (   31.46 Msamples/s)
    AVX2     :      0.1492 ms   (    3.35 Msamples/s)
    Speedup  :        0.11x

  ACCURACY (AVX2 vs scalar reference):
    Output         |    avg |err|    max |err| |  avg ULP    max ULP |      avg rel      max rel
    --------------------------------------------------------------------------------------------------
    V_re           |    9.918e-08    4.768e-07 |      2.1         11 |    1.660e-07    8.769e-07
    V_im           |    1.549e-07    5.960e-07 |      2.8         11 |    2.267e-07    7.428e-07
    H_re           |    2.629e-08    1.192e-07 |      1.8          8 |    1.609e-07    7.206e-07
    H_im           |    2.854e-08    1.192e-07 |      2.0          8 |    1.700e-07    5.900e-07
    dist           |    0.000e+00    0.000e+00 |      0.0          0 |    0.000e+00    0.000e+00
    az_loc         |    2.971e-08    1.192e-07 |      0.7          4 |    6.301e-08    2.586e-07
    el_loc         |    6.288e-09    5.960e-08 |      0.3          2 |    2.733e-08    2.212e-07
    gamma          |    1.333e-08    6.333e-08 |      2.6         17 |    2.079e-07    1.291e-06

  COMPLEX FIELD ACCURACY:
    field      avg amp err  max amp err   avg ph err   max ph err  avg ULP(re)  max ULP(re)
    ----------------------------------------------------------------------------------------
    V (theta)     1.547e-07    6.195e-07    1.119e-07    3.675e-07          2.1           11
    H (phi)       3.766e-08    1.488e-07    4.527e-08    1.682e-07          1.8            8

========================================================================
  S12: float  n_out=64  n_ang=500  grid=1deg (361x181)
  Pattern: 361 x 181 x 64   n_out: 64   n_ang: 500
  per_element_rotation: YES   dtype: float   n_warmup: 500   n_reps: 100
========================================================================

  PERFORMANCE (per call):
    Original :      0.6818 ms   (   46.93 Msamples/s)
    AVX2     :      0.0832 ms   (  384.78 Msamples/s)
    Speedup  :        8.20x

  ACCURACY (AVX2 vs scalar reference):
    Output         |    avg |err|    max |err| |  avg ULP    max ULP |      avg rel      max rel
    --------------------------------------------------------------------------------------------------
    V_re           |    6.941e-08    1.407e-05 |     27.2     409645 |    2.424e-06    3.858e-02
    V_im           |    7.475e-08    2.296e-05 |     20.3     124604 |    1.733e-06    1.153e-02
    H_re           |    8.382e-08    1.416e-05 |      6.5       6570 |    5.358e-07    6.858e-04
    H_im           |    8.550e-08    1.248e-05 |      9.2      25516 |    7.872e-07    3.030e-03
    dist           |    5.382e-07    5.722e-06 |      0.7          5 |    5.146e-08    3.548e-07
    az_loc         |    2.940e-08    2.384e-07 |      2.6      12841 |    2.132e-07    7.713e-04
    el_loc         |    3.273e-08    3.576e-07 |     13.1     314834 |    1.001e-06    2.414e-02
    gamma          |    5.638e-08    4.768e-07 |      0.6         17 |    4.896e-08    1.291e-06

  COMPLEX FIELD ACCURACY:
    field      avg amp err  max amp err   avg ph err   max ph err  avg ULP(re)  max ULP(re)
    ----------------------------------------------------------------------------------------
    V (theta)     8.332e-08    1.381e-05    1.637e-07    8.433e-05         27.2       409645
    H (phi)       1.018e-07    8.565e-06    9.987e-08    1.756e-05          6.5         6570

========================================================================
  S13: float  n_out=64  n_ang=2000 grid=1deg (361x181)
  Pattern: 361 x 181 x 64   n_out: 64   n_ang: 2000
  per_element_rotation: YES   dtype: float   n_warmup: 500   n_reps: 50
========================================================================

  PERFORMANCE (per call):
    Original :      2.6455 ms   (   48.38 Msamples/s)
    AVX2     :      0.2777 ms   (  460.86 Msamples/s)
    Speedup  :        9.52x

  ACCURACY (AVX2 vs scalar reference):
    Output         |    avg |err|    max |err| |  avg ULP    max ULP |      avg rel      max rel
    --------------------------------------------------------------------------------------------------
    V_re           |    8.387e-08    2.676e-05 |     20.8     409645 |    1.710e-06    3.858e-02
    V_im           |    8.529e-08    2.450e-05 |     15.1     124604 |    1.272e-06    1.153e-02
    H_re           |    7.800e-08    1.699e-05 |     26.7     947319 |    2.245e-06    8.799e-02
    H_im           |    7.851e-08    1.747e-05 |      7.7      93150 |    6.307e-07    7.756e-03
    dist           |    5.693e-07    5.722e-06 |      0.8       2478 |    6.867e-08    1.833e-04
    az_loc         |    3.088e-08    2.384e-07 |      1.7      12841 |    1.403e-07    7.713e-04
    el_loc         |    2.767e-08    3.576e-07 |      4.8     314834 |    3.793e-07    2.414e-02
    gamma          |    4.833e-08    4.768e-07 |      2.4      15737 |    1.992e-07    1.629e-03

  COMPLEX FIELD ACCURACY:
    field      avg amp err  max amp err   avg ph err   max ph err  avg ULP(re)  max ULP(re)
    ----------------------------------------------------------------------------------------
    V (theta)     9.256e-08    1.381e-05    1.504e-07    8.433e-05         20.8       409645
    H (phi)       9.519e-08    9.637e-06    9.723e-08    2.186e-05         26.7       947319

========================================================================
  S14: double n_out=64  n_ang=500  grid=5deg (73x37)
  Pattern: 73 x 37 x 64   n_out: 64   n_ang: 500
  per_element_rotation: YES   dtype: double   n_warmup: 1000   n_reps: 100
========================================================================

  PERFORMANCE (per call):
    Original :      0.7211 ms   (   44.37 Msamples/s)
    AVX2     :      0.0946 ms   (  338.28 Msamples/s)
    Speedup  :        7.62x

  ACCURACY (AVX2 vs scalar reference):
    Output         |    avg |err|    max |err| |  avg ULP    max ULP |      avg rel      max rel
    --------------------------------------------------------------------------------------------------
    V_re           |    7.958e-08    6.083e-07 |     13.7      36251 |    1.143e-06    3.241e-03
    V_im           |    9.144e-08    7.313e-07 |     18.8      42436 |    1.631e-06    4.269e-03
    H_re           |    9.429e-08    5.458e-07 |      8.8      18944 |    7.286e-07    1.529e-03
    H_im           |    9.651e-08    6.040e-07 |     14.6      83182 |    1.181e-06    5.830e-03
    dist           |    1.953e-06    7.391e-06 |      2.5          8 |    1.950e-07    5.853e-07
    az_loc         |    4.964e-08    2.830e-07 |      6.2      13080 |    5.196e-07    8.445e-04
    el_loc         |    1.094e-07    6.130e-07 |     10.7      36969 |    8.287e-07    2.452e-03
    gamma          |    6.726e-08    4.450e-07 |      0.6         12 |    5.582e-08    8.995e-07

  COMPLEX FIELD ACCURACY:
    field      avg amp err  max amp err   avg ph err   max ph err  avg ULP(re)  max ULP(re)
    ----------------------------------------------------------------------------------------
    V (theta)     1.121e-07    7.345e-07    1.645e-07    3.825e-06         13.7        36251
    H (phi)       1.019e-07    6.424e-07    1.496e-07    1.652e-06          8.8        18944

========================================================================
  S15: double n_out=64  n_ang=2000 grid=5deg (73x37)
  Pattern: 73 x 37 x 64   n_out: 64   n_ang: 2000
  per_element_rotation: YES   dtype: double   n_warmup: 500   n_reps: 50
========================================================================

  PERFORMANCE (per call):
    Original :      2.4864 ms   (   51.48 Msamples/s)
    AVX2     :      0.3083 ms   (  415.15 Msamples/s)
    Speedup  :        8.06x

  ACCURACY (AVX2 vs scalar reference):
    Output         |    avg |err|    max |err| |  avg ULP    max ULP |      avg rel      max rel
    --------------------------------------------------------------------------------------------------
    V_re           |    8.569e-08    6.443e-07 |     16.8     117245 |    1.371e-06    1.029e-02
    V_im           |    8.699e-08    7.313e-07 |     16.5     154776 |    1.424e-06    1.765e-02
    H_re           |    9.603e-08    6.103e-07 |     60.8    3422847 |    4.689e-06    2.406e-01
    H_im           |    9.571e-08    6.040e-07 |     12.8      85573 |    1.033e-06    5.830e-03
    dist           |    1.683e-06    1.023e-05 |      3.0      40675 |    2.541e-07    3.768e-03
    az_loc         |    6.868e-08    2.935e-07 |      6.5      21160 |    5.426e-07    1.366e-03
    el_loc         |    9.763e-08    6.130e-07 |     11.3     123339 |    9.240e-07    1.058e-02
    gamma          |    6.529e-08    4.450e-07 |      3.2      19907 |    2.692e-07    2.061e-03

  COMPLEX FIELD ACCURACY:
    field      avg amp err  max amp err   avg ph err   max ph err  avg ULP(re)  max ULP(re)
    ----------------------------------------------------------------------------------------
    V (theta)     1.068e-07    7.345e-07    1.569e-07    3.825e-06         16.8       117245
    H (phi)       9.677e-08    6.424e-07    1.641e-07    1.984e-06         60.8      3422847

================================================================
  Benchmark complete.
================================================================

