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

1.06
rc_pixop_set_u8

2.01
rc_pixop_not_u8

3.72
rc_pixop_flip_u8

3.08
rc_pixop_abs_u8

1.13
rc_pixop_addc_u8

1.51
rc_pixop_lerpc_u8

4.65
rc_pixop_lerpnc_u8

3.93
rc_pixop_add_u8

2.38
rc_pixop_avg_u8

2.37
rc_pixop_sub_u8

1.69
rc_pixop_subh_u8

2.40
rc_pixop_suba_u8

1.33
rc_pixop_lerp_u8

1.57
rc_pixop_lerpn_u8

1.96
rc_pixop_lerpi_u8

1.75
rc_pixop_norm_u8

1.29
rc_type_u8_to_bin

1.43
rc_thresh_gt_u8

1.26
rc_thresh_lt_u8

1.21
rc_thresh_gtlt_u8

1.31
rc_thresh_ltgt_u8

1.35
rc_thresh_gt_pixel_u8

1.30
rc_thresh_lt_pixel_u8

1.30
rc_thresh_gtlt_pixel_u8

1.21
rc_thresh_ltgt_pixel_u8

1.22
rc_stat_sum_bin

1.11
rc_stat_sum_u8

1.43
rc_stat_min_bin

1.05
rc_stat_max_bin

1.02
rc_stat_min_u8

1.35
rc_stat_max_u8

1.48
rc_reduce_1x2_u8

1.01
rc_reduce_2x1_u8

2.81
rc_reduce_2x2_u8

1.06
rc_filter_diff_1x2_horz_u8

1.54
rc_filter_diff_1x2_horz_abs_u8

1.03
rc_filter_diff_2x1_vert_u8

2.24
rc_filter_diff_2x1_vert_abs_u8

1.40
rc_filter_diff_2x2_magn_u8

1.08
rc_filter_sobel_3x3_horz_u8

1.50
rc_filter_sobel_3x3_horz_abs_u8

1.42
rc_filter_sobel_3x3_vert_u8

1.01
rc_filter_sobel_3x3_vert_abs_u8

1.32
rc_filter_sobel_3x3_magn_u8

1.06
rc_filter_gauss_3x3_u8

1.41
rc_filter_laplace_3x3_u8

1.24
rc_filter_laplace_3x3_abs_u8

1.19
rc_filter_highpass_3x3_u8

1.98
rc_filter_highpass_3x3_abs_u8

1.81
rc_margin_horz_bin

1.01