Skip to content

x86: Add high bitdepth mc SSSE3 asm

Henrik Gramner requested to merge gramner/dav1d:mc16_ssse3 into master

Includes x86-32 support.

avg_w4_16bpc_c: 174.5
avg_w4_16bpc_ssse3: 16.9
avg_w4_16bpc_avx2: 15.3
avg_w8_16bpc_c: 243.4
avg_w8_16bpc_ssse3: 42.9
avg_w8_16bpc_avx2: 25.4
avg_w16_16bpc_c: 655.2
avg_w16_16bpc_ssse3: 140.1
avg_w16_16bpc_avx2: 55.2
avg_w32_16bpc_c: 2342.5
avg_w32_16bpc_ssse3: 464.9
avg_w32_16bpc_avx2: 211.9
avg_w64_16bpc_c: 5386.7
avg_w64_16bpc_ssse3: 1269.6
avg_w64_16bpc_avx2: 705.0
avg_w128_16bpc_c: 14241.7
avg_w128_16bpc_ssse3: 3566.7
avg_w128_16bpc_avx2: 2073.8

blend_h_w2_16bpc_c: 105.5
blend_h_w2_16bpc_ssse3: 21.5
blend_h_w2_16bpc_avx2: 22.9
blend_h_w4_16bpc_c: 148.3
blend_h_w4_16bpc_ssse3: 19.3
blend_h_w4_16bpc_avx2: 19.0
blend_h_w8_16bpc_c: 176.7
blend_h_w8_16bpc_ssse3: 23.3
blend_h_w8_16bpc_avx2: 22.0
blend_h_w16_16bpc_c: 319.3
blend_h_w16_16bpc_ssse3: 47.0
blend_h_w16_16bpc_avx2: 22.6
blend_h_w32_16bpc_c: 611.4
blend_h_w32_16bpc_ssse3: 62.4
blend_h_w32_16bpc_avx2: 38.8
blend_h_w64_16bpc_c: 1195.8
blend_h_w64_16bpc_ssse3: 103.9
blend_h_w64_16bpc_avx2: 67.0
blend_h_w128_16bpc_c: 2981.4
blend_h_w128_16bpc_ssse3: 222.6
blend_h_w128_16bpc_avx2: 117.2

blend_v_w2_16bpc_c: 152.9
blend_v_w2_16bpc_ssse3: 46.4
blend_v_w2_16bpc_avx2: 40.1
blend_v_w4_16bpc_c: 420.5
blend_v_w4_16bpc_ssse3: 50.9
blend_v_w4_16bpc_avx2: 50.4
blend_v_w8_16bpc_c: 667.3
blend_v_w8_16bpc_ssse3: 71.3
blend_v_w8_16bpc_avx2: 52.0
blend_v_w16_16bpc_c: 1203.6
blend_v_w16_16bpc_ssse3: 120.0
blend_v_w16_16bpc_avx2: 59.7
blend_v_w32_16bpc_c: 1854.3
blend_v_w32_16bpc_ssse3: 176.1
blend_v_w32_16bpc_avx2: 112.1

blend_w4_16bpc_c: 126.8
blend_w4_16bpc_ssse3: 17.2
blend_w4_16bpc_avx2: 18.7
blend_w8_16bpc_c: 293.4
blend_w8_16bpc_ssse3: 26.0
blend_w8_16bpc_avx2: 19.9
blend_w16_16bpc_c: 698.6
blend_w16_16bpc_ssse3: 85.9
blend_w16_16bpc_avx2: 53.4
blend_w32_16bpc_c: 1765.2
blend_w32_16bpc_ssse3: 206.9
blend_w32_16bpc_avx2: 109.0

mask_w4_16bpc_c: 234.6
mask_w4_16bpc_ssse3: 34.8
mask_w4_16bpc_avx2: 36.1
mask_w8_16bpc_c: 624.7
mask_w8_16bpc_ssse3: 86.2
mask_w8_16bpc_avx2: 55.7
mask_w16_16bpc_c: 920.0
mask_w16_16bpc_ssse3: 269.9
mask_w16_16bpc_avx2: 141.4
mask_w32_16bpc_c: 3433.9
mask_w32_16bpc_ssse3: 1057.0
mask_w32_16bpc_avx2: 520.0
mask_w64_16bpc_c: 8108.8
mask_w64_16bpc_ssse3: 2411.4
mask_w64_16bpc_avx2: 1295.9
mask_w128_16bpc_c: 19880.3
mask_w128_16bpc_ssse3: 6031.1
mask_w128_16bpc_avx2: 3246.5

mc_8tap_regular_w2_0_16bpc_c: 103.0
mc_8tap_regular_w2_0_16bpc_ssse3: 17.0
mc_8tap_regular_w2_0_16bpc_avx2: 17.1
mc_8tap_regular_w2_h_16bpc_c: 301.6
mc_8tap_regular_w2_h_16bpc_ssse3: 44.6
mc_8tap_regular_w2_h_16bpc_avx2: 38.8
mc_8tap_regular_w2_hv_16bpc_c: 693.9
mc_8tap_regular_w2_hv_16bpc_ssse3: 89.5
mc_8tap_regular_w2_hv_16bpc_avx2: 83.8
mc_8tap_regular_w2_v_16bpc_c: 483.7
mc_8tap_regular_w2_v_16bpc_ssse3: 57.9
mc_8tap_regular_w2_v_16bpc_avx2: 50.2
mc_8tap_regular_w4_0_16bpc_c: 110.8
mc_8tap_regular_w4_0_16bpc_ssse3: 16.2
mc_8tap_regular_w4_0_16bpc_avx2: 17.4
mc_8tap_regular_w4_h_16bpc_c: 519.8
mc_8tap_regular_w4_h_16bpc_ssse3: 63.2
mc_8tap_regular_w4_h_16bpc_avx2: 41.9
mc_8tap_regular_w4_hv_16bpc_c: 1207.0
mc_8tap_regular_w4_hv_16bpc_ssse3: 166.7
mc_8tap_regular_w4_hv_16bpc_avx2: 87.8
mc_8tap_regular_w4_v_16bpc_c: 710.7
mc_8tap_regular_w4_v_16bpc_ssse3: 65.6
mc_8tap_regular_w4_v_16bpc_avx2: 50.9
mc_8tap_regular_w8_0_16bpc_c: 103.2
mc_8tap_regular_w8_0_16bpc_ssse3: 16.6
mc_8tap_regular_w8_0_16bpc_avx2: 16.6
mc_8tap_regular_w8_h_16bpc_c: 955.9
mc_8tap_regular_w8_h_16bpc_ssse3: 131.4
mc_8tap_regular_w8_h_16bpc_avx2: 80.4
mc_8tap_regular_w8_hv_16bpc_c: 2070.0
mc_8tap_regular_w8_hv_16bpc_ssse3: 315.5
mc_8tap_regular_w8_hv_16bpc_avx2: 202.7
mc_8tap_regular_w8_v_16bpc_c: 1134.6
mc_8tap_regular_w8_v_16bpc_ssse3: 119.7
mc_8tap_regular_w8_v_16bpc_avx2: 74.9
mc_8tap_regular_w16_0_16bpc_c: 128.0
mc_8tap_regular_w16_0_16bpc_ssse3: 33.3
mc_8tap_regular_w16_0_16bpc_avx2: 24.2
mc_8tap_regular_w16_h_16bpc_c: 3055.8
mc_8tap_regular_w16_h_16bpc_ssse3: 335.5
mc_8tap_regular_w16_h_16bpc_avx2: 181.8
mc_8tap_regular_w16_hv_16bpc_c: 4249.8
mc_8tap_regular_w16_hv_16bpc_ssse3: 815.5
mc_8tap_regular_w16_hv_16bpc_avx2: 518.7
mc_8tap_regular_w16_v_16bpc_c: 3356.3
mc_8tap_regular_w16_v_16bpc_ssse3: 318.0
mc_8tap_regular_w16_v_16bpc_avx2: 185.3
mc_8tap_regular_w32_0_16bpc_c: 193.2
mc_8tap_regular_w32_0_16bpc_ssse3: 110.7
mc_8tap_regular_w32_0_16bpc_avx2: 66.6
mc_8tap_regular_w32_h_16bpc_c: 8931.5
mc_8tap_regular_w32_h_16bpc_ssse3: 1053.6
mc_8tap_regular_w32_h_16bpc_avx2: 543.9
mc_8tap_regular_w32_hv_16bpc_c: 11748.1
mc_8tap_regular_w32_hv_16bpc_ssse3: 2388.0
mc_8tap_regular_w32_hv_16bpc_avx2: 1525.5
mc_8tap_regular_w32_v_16bpc_c: 9407.3
mc_8tap_regular_w32_v_16bpc_ssse3: 973.2
mc_8tap_regular_w32_v_16bpc_avx2: 546.9
mc_8tap_regular_w64_0_16bpc_c: 695.0
mc_8tap_regular_w64_0_16bpc_ssse3: 489.7
mc_8tap_regular_w64_0_16bpc_avx2: 358.4
mc_8tap_regular_w64_h_16bpc_c: 30537.3
mc_8tap_regular_w64_h_16bpc_ssse3: 3606.7
mc_8tap_regular_w64_h_16bpc_avx2: 1984.0
mc_8tap_regular_w64_hv_16bpc_c: 37932.8
mc_8tap_regular_w64_hv_16bpc_ssse3: 8200.4
mc_8tap_regular_w64_hv_16bpc_avx2: 5190.5
mc_8tap_regular_w64_v_16bpc_c: 31382.8
mc_8tap_regular_w64_v_16bpc_ssse3: 3539.2
mc_8tap_regular_w64_v_16bpc_avx2: 1999.4
mc_8tap_regular_w128_0_16bpc_c: 1311.0
mc_8tap_regular_w128_0_16bpc_ssse3: 1393.1
mc_8tap_regular_w128_0_16bpc_avx2: 992.8
mc_8tap_regular_w128_h_16bpc_c: 84498.5
mc_8tap_regular_w128_h_16bpc_ssse3: 9980.3
mc_8tap_regular_w128_h_16bpc_avx2: 5555.8
mc_8tap_regular_w128_hv_16bpc_c: 102460.9
mc_8tap_regular_w128_hv_16bpc_ssse3: 23405.2
mc_8tap_regular_w128_hv_16bpc_avx2: 14830.1
mc_8tap_regular_w128_v_16bpc_c: 85728.1
mc_8tap_regular_w128_v_16bpc_ssse3: 10806.1
mc_8tap_regular_w128_v_16bpc_avx2: 6407.0

mc_bilinear_w2_0_16bpc_c: 96.6
mc_bilinear_w2_0_16bpc_ssse3: 15.5
mc_bilinear_w2_0_16bpc_avx2: 16.1
mc_bilinear_w2_h_16bpc_c: 201.6
mc_bilinear_w2_h_16bpc_ssse3: 31.5
mc_bilinear_w2_h_16bpc_avx2: 28.8
mc_bilinear_w2_hv_16bpc_c: 263.1
mc_bilinear_w2_hv_16bpc_ssse3: 47.6
mc_bilinear_w2_hv_16bpc_avx2: 48.1
mc_bilinear_w2_v_16bpc_c: 171.8
mc_bilinear_w2_v_16bpc_ssse3: 25.8
mc_bilinear_w2_v_16bpc_avx2: 26.5
mc_bilinear_w4_0_16bpc_c: 106.0
mc_bilinear_w4_0_16bpc_ssse3: 15.1
mc_bilinear_w4_0_16bpc_avx2: 15.9
mc_bilinear_w4_h_16bpc_c: 323.8
mc_bilinear_w4_h_16bpc_ssse3: 31.8
mc_bilinear_w4_h_16bpc_avx2: 30.2
mc_bilinear_w4_hv_16bpc_c: 428.7
mc_bilinear_w4_hv_16bpc_ssse3: 45.8
mc_bilinear_w4_hv_16bpc_avx2: 46.4
mc_bilinear_w4_v_16bpc_c: 251.5
mc_bilinear_w4_v_16bpc_ssse3: 26.3
mc_bilinear_w4_v_16bpc_avx2: 24.4
mc_bilinear_w8_0_16bpc_c: 96.2
mc_bilinear_w8_0_16bpc_ssse3: 15.2
mc_bilinear_w8_0_16bpc_avx2: 15.3
mc_bilinear_w8_h_16bpc_c: 280.1
mc_bilinear_w8_h_16bpc_ssse3: 32.5
mc_bilinear_w8_h_16bpc_avx2: 32.2
mc_bilinear_w8_hv_16bpc_c: 428.0
mc_bilinear_w8_hv_16bpc_ssse3: 60.7
mc_bilinear_w8_hv_16bpc_avx2: 51.2
mc_bilinear_w8_v_16bpc_c: 258.5
mc_bilinear_w8_v_16bpc_ssse3: 31.2
mc_bilinear_w8_v_16bpc_avx2: 28.4
mc_bilinear_w16_0_16bpc_c: 121.3
mc_bilinear_w16_0_16bpc_ssse3: 32.5
mc_bilinear_w16_0_16bpc_avx2: 23.1
mc_bilinear_w16_h_16bpc_c: 716.0
mc_bilinear_w16_h_16bpc_ssse3: 97.0
mc_bilinear_w16_h_16bpc_avx2: 48.9
mc_bilinear_w16_hv_16bpc_c: 1159.7
mc_bilinear_w16_hv_16bpc_ssse3: 155.2
mc_bilinear_w16_hv_16bpc_avx2: 89.2
mc_bilinear_w16_v_16bpc_c: 644.8
mc_bilinear_w16_v_16bpc_ssse3: 78.7
mc_bilinear_w16_v_16bpc_avx2: 45.6
mc_bilinear_w32_0_16bpc_c: 188.3
mc_bilinear_w32_0_16bpc_ssse3: 109.9
mc_bilinear_w32_0_16bpc_avx2: 68.6
mc_bilinear_w32_h_16bpc_c: 2221.7
mc_bilinear_w32_h_16bpc_ssse3: 251.8
mc_bilinear_w32_h_16bpc_avx2: 131.3
mc_bilinear_w32_hv_16bpc_c: 3615.8
mc_bilinear_w32_hv_16bpc_ssse3: 472.4
mc_bilinear_w32_hv_16bpc_avx2: 266.2
mc_bilinear_w32_v_16bpc_c: 1957.8
mc_bilinear_w32_v_16bpc_ssse3: 228.2
mc_bilinear_w32_v_16bpc_avx2: 98.1
mc_bilinear_w64_0_16bpc_c: 657.5
mc_bilinear_w64_0_16bpc_ssse3: 485.8
mc_bilinear_w64_0_16bpc_avx2: 353.3
mc_bilinear_w64_h_16bpc_c: 7671.0
mc_bilinear_w64_h_16bpc_ssse3: 942.0
mc_bilinear_w64_h_16bpc_avx2: 565.7
mc_bilinear_w64_hv_16bpc_c: 12646.2
mc_bilinear_w64_hv_16bpc_ssse3: 1758.6
mc_bilinear_w64_hv_16bpc_avx2: 971.0
mc_bilinear_w64_v_16bpc_c: 6778.3
mc_bilinear_w64_v_16bpc_ssse3: 867.9
mc_bilinear_w64_v_16bpc_avx2: 523.3
mc_bilinear_w128_0_16bpc_c: 1340.6
mc_bilinear_w128_0_16bpc_ssse3: 1380.6
mc_bilinear_w128_0_16bpc_avx2: 978.3
mc_bilinear_w128_h_16bpc_c: 21285.5
mc_bilinear_w128_h_16bpc_ssse3: 2407.2
mc_bilinear_w128_h_16bpc_avx2: 1534.3
mc_bilinear_w128_hv_16bpc_c: 35713.0
mc_bilinear_w128_hv_16bpc_ssse3: 5384.3
mc_bilinear_w128_hv_16bpc_avx2: 3007.0
mc_bilinear_w128_v_16bpc_c: 18738.1
mc_bilinear_w128_v_16bpc_ssse3: 3411.4
mc_bilinear_w128_v_16bpc_avx2: 1854.1

mct_8tap_regular_w4_0_16bpc_c: 182.9
mct_8tap_regular_w4_0_16bpc_ssse3: 15.0
mct_8tap_regular_w4_0_16bpc_avx2: 16.5
mct_8tap_regular_w4_h_16bpc_c: 386.7
mct_8tap_regular_w4_h_16bpc_ssse3: 34.3
mct_8tap_regular_w4_h_16bpc_avx2: 24.4
mct_8tap_regular_w4_hv_16bpc_c: 947.8
mct_8tap_regular_w4_hv_16bpc_ssse3: 126.4
mct_8tap_regular_w4_hv_16bpc_avx2: 65.3
mct_8tap_regular_w4_v_16bpc_c: 523.5
mct_8tap_regular_w4_v_16bpc_ssse3: 53.3
mct_8tap_regular_w4_v_16bpc_avx2: 38.4
mct_8tap_regular_w8_0_16bpc_c: 135.4
mct_8tap_regular_w8_0_16bpc_ssse3: 22.7
mct_8tap_regular_w8_0_16bpc_avx2: 20.3
mct_8tap_regular_w8_h_16bpc_c: 1117.2
mct_8tap_regular_w8_h_16bpc_ssse3: 132.9
mct_8tap_regular_w8_h_16bpc_avx2: 86.6
mct_8tap_regular_w8_hv_16bpc_c: 2296.9
mct_8tap_regular_w8_hv_16bpc_ssse3: 331.3
mct_8tap_regular_w8_hv_16bpc_avx2: 216.8
mct_8tap_regular_w8_v_16bpc_c: 1278.0
mct_8tap_regular_w8_v_16bpc_ssse3: 148.6
mct_8tap_regular_w8_v_16bpc_avx2: 81.3
mct_8tap_regular_w16_0_16bpc_c: 359.8
mct_8tap_regular_w16_0_16bpc_ssse3: 65.0
mct_8tap_regular_w16_0_16bpc_avx2: 36.9
mct_8tap_regular_w16_h_16bpc_c: 4054.9
mct_8tap_regular_w16_h_16bpc_ssse3: 388.4
mct_8tap_regular_w16_h_16bpc_avx2: 222.6
mct_8tap_regular_w16_hv_16bpc_c: 5175.9
mct_8tap_regular_w16_hv_16bpc_ssse3: 980.7
mct_8tap_regular_w16_hv_16bpc_avx2: 633.0
mct_8tap_regular_w16_v_16bpc_c: 4428.0
mct_8tap_regular_w16_v_16bpc_ssse3: 447.2
mct_8tap_regular_w16_v_16bpc_avx2: 228.4
mct_8tap_regular_w32_0_16bpc_c: 1365.7
mct_8tap_regular_w32_0_16bpc_ssse3: 214.6
mct_8tap_regular_w32_0_16bpc_avx2: 127.3
mct_8tap_regular_w32_h_16bpc_c: 15070.0
mct_8tap_regular_w32_h_16bpc_ssse3: 1620.7
mct_8tap_regular_w32_h_16bpc_avx2: 846.9
mct_8tap_regular_w32_hv_16bpc_c: 17941.8
mct_8tap_regular_w32_hv_16bpc_ssse3: 3625.8
mct_8tap_regular_w32_hv_16bpc_avx2: 2341.2
mct_8tap_regular_w32_v_16bpc_c: 15799.4
mct_8tap_regular_w32_v_16bpc_ssse3: 1677.8
mct_8tap_regular_w32_v_16bpc_avx2: 854.2
mct_8tap_regular_w64_0_16bpc_c: 3277.8
mct_8tap_regular_w64_0_16bpc_ssse3: 582.3
mct_8tap_regular_w64_0_16bpc_avx2: 444.0
mct_8tap_regular_w64_h_16bpc_c: 35112.7
mct_8tap_regular_w64_h_16bpc_ssse3: 3757.8
mct_8tap_regular_w64_h_16bpc_avx2: 2192.7
mct_8tap_regular_w64_hv_16bpc_c: 40844.1
mct_8tap_regular_w64_hv_16bpc_ssse3: 8834.9
mct_8tap_regular_w64_hv_16bpc_avx2: 5686.4
mct_8tap_regular_w64_v_16bpc_c: 35995.7
mct_8tap_regular_w64_v_16bpc_ssse3: 4142.8
mct_8tap_regular_w64_v_16bpc_avx2: 2217.6
mct_8tap_regular_w128_0_16bpc_c: 8171.2
mct_8tap_regular_w128_0_16bpc_ssse3: 1349.9
mct_8tap_regular_w128_0_16bpc_avx2: 1037.3
mct_8tap_regular_w128_h_16bpc_c: 85723.4
mct_8tap_regular_w128_h_16bpc_ssse3: 9077.4
mct_8tap_regular_w128_h_16bpc_avx2: 5228.1
mct_8tap_regular_w128_hv_16bpc_c: 98040.4
mct_8tap_regular_w128_hv_16bpc_ssse3: 22341.9
mct_8tap_regular_w128_hv_16bpc_avx2: 14246.6
mct_8tap_regular_w128_v_16bpc_c: 86804.5
mct_8tap_regular_w128_v_16bpc_ssse3: 10809.3
mct_8tap_regular_w128_v_16bpc_avx2: 5982.8

mct_bilinear_w4_0_16bpc_c: 132.4
mct_bilinear_w4_0_16bpc_ssse3: 12.8
mct_bilinear_w4_0_16bpc_avx2: 14.3
mct_bilinear_w4_h_16bpc_c: 174.5
mct_bilinear_w4_h_16bpc_ssse3: 21.1
mct_bilinear_w4_h_16bpc_avx2: 18.5
mct_bilinear_w4_hv_16bpc_c: 252.0
mct_bilinear_w4_hv_16bpc_ssse3: 29.1
mct_bilinear_w4_hv_16bpc_avx2: 26.5
mct_bilinear_w4_v_16bpc_c: 208.0
mct_bilinear_w4_v_16bpc_ssse3: 22.5
mct_bilinear_w4_v_16bpc_avx2: 17.6
mct_bilinear_w8_0_16bpc_c: 131.8
mct_bilinear_w8_0_16bpc_ssse3: 21.5
mct_bilinear_w8_0_16bpc_avx2: 20.4
mct_bilinear_w8_h_16bpc_c: 228.5
mct_bilinear_w8_h_16bpc_ssse3: 40.2
mct_bilinear_w8_h_16bpc_avx2: 32.5
mct_bilinear_w8_hv_16bpc_c: 414.0
mct_bilinear_w8_hv_16bpc_ssse3: 58.3
mct_bilinear_w8_hv_16bpc_avx2: 44.2
mct_bilinear_w8_v_16bpc_c: 245.8
mct_bilinear_w8_v_16bpc_ssse3: 42.9
mct_bilinear_w8_v_16bpc_avx2: 32.5
mct_bilinear_w16_0_16bpc_c: 356.0
mct_bilinear_w16_0_16bpc_ssse3: 62.1
mct_bilinear_w16_0_16bpc_avx2: 37.0
mct_bilinear_w16_h_16bpc_c: 667.1
mct_bilinear_w16_h_16bpc_ssse3: 128.8
mct_bilinear_w16_h_16bpc_avx2: 61.1
mct_bilinear_w16_hv_16bpc_c: 1317.5
mct_bilinear_w16_hv_16bpc_ssse3: 163.9
mct_bilinear_w16_hv_16bpc_avx2: 93.9
mct_bilinear_w16_v_16bpc_c: 682.8
mct_bilinear_w16_v_16bpc_ssse3: 109.8
mct_bilinear_w16_v_16bpc_avx2: 63.3
mct_bilinear_w32_0_16bpc_c: 1360.8
mct_bilinear_w32_0_16bpc_ssse3: 213.2
mct_bilinear_w32_0_16bpc_avx2: 132.6
mct_bilinear_w32_h_16bpc_c: 2628.3
mct_bilinear_w32_h_16bpc_ssse3: 418.4
mct_bilinear_w32_h_16bpc_avx2: 255.6
mct_bilinear_w32_hv_16bpc_c: 5198.1
mct_bilinear_w32_hv_16bpc_ssse3: 628.3
mct_bilinear_w32_hv_16bpc_avx2: 346.3
mct_bilinear_w32_v_16bpc_c: 2639.9
mct_bilinear_w32_v_16bpc_ssse3: 418.4
mct_bilinear_w32_v_16bpc_avx2: 232.8
mct_bilinear_w64_0_16bpc_c: 3272.9
mct_bilinear_w64_0_16bpc_ssse3: 581.1
mct_bilinear_w64_0_16bpc_avx2: 446.1
mct_bilinear_w64_h_16bpc_c: 6259.6
mct_bilinear_w64_h_16bpc_ssse3: 1784.8
mct_bilinear_w64_h_16bpc_avx2: 647.1
mct_bilinear_w64_hv_16bpc_c: 12481.7
mct_bilinear_w64_hv_16bpc_ssse3: 1615.1
mct_bilinear_w64_hv_16bpc_avx2: 922.4
mct_bilinear_w64_v_16bpc_c: 6298.3
mct_bilinear_w64_v_16bpc_ssse3: 1103.1
mct_bilinear_w64_v_16bpc_avx2: 649.0
mct_bilinear_w128_0_16bpc_c: 8109.9
mct_bilinear_w128_0_16bpc_ssse3: 1380.0
mct_bilinear_w128_0_16bpc_avx2: 1024.8
mct_bilinear_w128_h_16bpc_c: 16001.4
mct_bilinear_w128_h_16bpc_ssse3: 3712.1
mct_bilinear_w128_h_16bpc_avx2: 1551.6
mct_bilinear_w128_hv_16bpc_c: 31935.9
mct_bilinear_w128_hv_16bpc_ssse3: 4380.3
mct_bilinear_w128_hv_16bpc_avx2: 2550.7
mct_bilinear_w128_v_16bpc_c: 16152.5
mct_bilinear_w128_v_16bpc_ssse3: 3466.8
mct_bilinear_w128_v_16bpc_avx2: 2000.0

w_avg_w4_16bpc_c: 200.7
w_avg_w4_16bpc_ssse3: 28.5
w_avg_w4_16bpc_avx2: 24.6
w_avg_w8_16bpc_c: 360.8
w_avg_w8_16bpc_ssse3: 65.1
w_avg_w8_16bpc_avx2: 39.9
w_avg_w16_16bpc_c: 1057.9
w_avg_w16_16bpc_ssse3: 215.5
w_avg_w16_16bpc_avx2: 91.0
w_avg_w32_16bpc_c: 3958.8
w_avg_w32_16bpc_ssse3: 723.0
w_avg_w32_16bpc_avx2: 332.6
w_avg_w64_16bpc_c: 9308.2
w_avg_w64_16bpc_ssse3: 1781.0
w_avg_w64_16bpc_avx2: 877.9
w_avg_w128_16bpc_c: 22828.0
w_avg_w128_16bpc_ssse3: 4353.0
w_avg_w128_16bpc_avx2: 2340.7

w_mask_420_w4_16bpc_c: 313.4
w_mask_420_w4_16bpc_ssse3: 43.9
w_mask_420_w4_16bpc_avx2: 37.1
w_mask_420_w8_16bpc_c: 954.8
w_mask_420_w8_16bpc_ssse3: 112.2
w_mask_420_w8_16bpc_avx2: 64.6
w_mask_420_w16_16bpc_c: 3106.4
w_mask_420_w16_16bpc_ssse3: 329.1
w_mask_420_w16_16bpc_avx2: 176.0
w_mask_420_w32_16bpc_c: 12275.3
w_mask_420_w32_16bpc_ssse3: 1285.3
w_mask_420_w32_16bpc_avx2: 631.0
w_mask_420_w64_16bpc_c: 29694.4
w_mask_420_w64_16bpc_ssse3: 3049.6
w_mask_420_w64_16bpc_avx2: 1582.3
w_mask_420_w128_16bpc_c: 74075.7
w_mask_420_w128_16bpc_ssse3: 7539.1
w_mask_420_w128_16bpc_avx2: 3964.5

w_mask_422_w4_16bpc_c: 309.4
w_mask_422_w4_16bpc_ssse3: 42.0
w_mask_422_w4_16bpc_avx2: 36.3
w_mask_422_w8_16bpc_c: 956.0
w_mask_422_w8_16bpc_ssse3: 115.3
w_mask_422_w8_16bpc_avx2: 63.1
w_mask_422_w16_16bpc_c: 3069.8
w_mask_422_w16_16bpc_ssse3: 341.0
w_mask_422_w16_16bpc_avx2: 168.2
w_mask_422_w32_16bpc_c: 12133.0
w_mask_422_w32_16bpc_ssse3: 1359.1
w_mask_422_w32_16bpc_avx2: 642.2
w_mask_422_w64_16bpc_c: 29842.9
w_mask_422_w64_16bpc_ssse3: 3229.7
w_mask_422_w64_16bpc_avx2: 1606.1
w_mask_422_w128_16bpc_c: 73382.0
w_mask_422_w128_16bpc_ssse3: 7927.5
w_mask_422_w128_16bpc_avx2: 4096.5

w_mask_444_w4_16bpc_c: 414.7
w_mask_444_w4_16bpc_ssse3: 39.8
w_mask_444_w4_16bpc_avx2: 28.2
w_mask_444_w8_16bpc_c: 1129.7
w_mask_444_w8_16bpc_ssse3: 108.5
w_mask_444_w8_16bpc_avx2: 59.7
w_mask_444_w16_16bpc_c: 1535.4
w_mask_444_w16_16bpc_ssse3: 317.3
w_mask_444_w16_16bpc_avx2: 164.8
w_mask_444_w32_16bpc_c: 5995.9
w_mask_444_w32_16bpc_ssse3: 1283.5
w_mask_444_w32_16bpc_avx2: 621.7
w_mask_444_w64_16bpc_c: 14506.5
w_mask_444_w64_16bpc_ssse3: 3044.5
w_mask_444_w64_16bpc_avx2: 1558.3
w_mask_444_w128_16bpc_c: 36094.8
w_mask_444_w128_16bpc_ssse3: 7402.8
w_mask_444_w128_16bpc_avx2: 3991.5

Merge request reports