prep_8tap avx512icl

• H
• V
• HV

mct_8tap_regular_w4_h_8bpc_avx2: 4.7
mct_8tap_regular_w4_h_8bpc_avx512icl: 4.3
mct_8tap_regular_w8_h_8bpc_avx2: 11.7
mct_8tap_regular_w8_h_8bpc_avx512icl: 7.7
mct_8tap_regular_w16_h_8bpc_avx2: 32.7
mct_8tap_regular_w16_h_8bpc_avx512icl: 16.2
mct_8tap_regular_w32_h_8bpc_avx2: 125.3
mct_8tap_regular_w32_h_8bpc_avx512icl: 60.3
mct_8tap_regular_w64_h_8bpc_avx2: 279.1
mct_8tap_regular_w64_h_8bpc_avx512icl: 137.4
mct_8tap_regular_w128_h_8bpc_avx2: 697.2
mct_8tap_regular_w128_h_8bpc_avx512icl: 341.2

mct_8tap_regular_w4_v_8bpc_avx2: 7.1
mct_8tap_regular_w4_v_8bpc_avx512icl: 7.3
mct_8tap_regular_w8_v_8bpc_avx2: 11.0
mct_8tap_regular_w8_v_8bpc_avx512icl: 9.9
mct_8tap_regular_w16_v_8bpc_avx2: 27.0
mct_8tap_regular_w16_v_8bpc_avx512icl: 20.1
mct_8tap_regular_w32_v_8bpc_avx2: 95.4
mct_8tap_regular_w32_v_8bpc_avx512icl: 61.7
mct_8tap_regular_w64_v_8bpc_avx2: 224.8
mct_8tap_regular_w64_v_8bpc_avx512icl: 141.0
mct_8tap_regular_w128_v_8bpc_avx2: 567.6
mct_8tap_regular_w128_v_8bpc_avx512icl: 346.1

mct_8tap_regular_w4_hv_8bpc_avx2: 13.5
mct_8tap_regular_w4_hv_8bpc_avx512icl: 9.9
mct_8tap_regular_w8_hv_8bpc_avx2: 38.8
mct_8tap_regular_w8_hv_8bpc_avx512icl: 22.1
mct_8tap_regular_w16_hv_8bpc_avx2: 112.9
mct_8tap_regular_w16_hv_8bpc_avx512icl: 49.7
mct_8tap_regular_w32_hv_8bpc_avx2: 414.1
mct_8tap_regular_w32_hv_8bpc_avx512icl: 178.2
mct_8tap_regular_w64_hv_8bpc_avx2: 987.4
mct_8tap_regular_w64_hv_8bpc_avx512icl: 422.8
mct_8tap_regular_w128_hv_8bpc_avx2: 2457.2
mct_8tap_regular_w128_hv_8bpc_avx512icl: 1042.7