RAPP Compute Tuning Report


Best generic

Best SWAR

Best SIMD
Linux gcc1-power7.osuosl.org 3.17.2-200.fc20.ppc64p7 #1 SMP Wed Feb 18 20:59:33 MST 2015 ppc64 ppc64 ppc64 GNU/Linux
Function
0%Relative Execution Speed100%
Gain
rc_bitblt_wa_copy_bin

1.07
rc_pixop_set_u8

1.38
rc_pixop_not_u8

7.15
rc_pixop_flip_u8

7.11
rc_pixop_abs_u8

1.71
rc_pixop_addc_u8

2.82
rc_pixop_lerpc_u8

8.66
rc_pixop_lerpnc_u8

7.42
rc_pixop_add_u8

4.60
rc_pixop_avg_u8

4.37
rc_pixop_sub_u8

3.24
rc_pixop_subh_u8

4.49
rc_pixop_suba_u8

2.75
rc_pixop_lerp_u8

3.55
rc_pixop_lerpn_u8

4.05
rc_pixop_lerpi_u8

3.85
rc_pixop_norm_u8

1.52
rc_type_u8_to_bin

2.95
rc_type_bin_to_u8

2.13
rc_thresh_gt_u8

3.30
rc_thresh_lt_u8

3.14
rc_thresh_gtlt_u8

3.38
rc_thresh_ltgt_u8

3.35
rc_thresh_gt_pixel_u8

3.50
rc_thresh_lt_pixel_u8

3.50
rc_thresh_gtlt_pixel_u8

3.23
rc_thresh_ltgt_pixel_u8

3.24
rc_stat_sum_bin

1.42
rc_stat_sum_u8

2.76
rc_stat_min_bin

1.06
rc_stat_max_bin

1.16
rc_stat_min_u8

1.43
rc_stat_max_u8

1.49
rc_reduce_1x2_u8

1.43
rc_reduce_2x1_u8

4.81
rc_reduce_2x2_u8

1.49
rc_filter_diff_1x2_horz_u8

2.91
rc_filter_diff_1x2_horz_abs_u8

2.01
rc_filter_diff_2x1_vert_u8

4.19
rc_filter_diff_2x1_vert_abs_u8

2.67
rc_filter_diff_2x2_magn_u8

2.35
rc_filter_sobel_3x3_horz_u8

3.06
rc_filter_sobel_3x3_horz_abs_u8

3.03
rc_filter_sobel_3x3_vert_u8

2.14
rc_filter_sobel_3x3_vert_abs_u8

2.83
rc_filter_sobel_3x3_magn_u8

2.13
rc_filter_gauss_3x3_u8

2.91
rc_filter_laplace_3x3_u8

2.61
rc_filter_laplace_3x3_abs_u8

2.21
rc_filter_highpass_3x3_u8

1.46
rc_filter_highpass_3x3_abs_u8

1.60
rc_margin_horz_bin

1.03
rc_cond_set_u8

1.98
rc_cond_addc_u8

1.13
rc_cond_subc_u8

1.44
rc_cond_copy_u8

2.07
rc_cond_add_u8

1.15