x86: Add high bit-depth ipred z3 AVX-512 (Ice Lake) asm

                                         Ice Lake          Zen 4

intra_pred_z3_w4_16bpc_avx2:           30.1 ( 6.07x)    25.0 ( 6.35x)
intra_pred_z3_w4_16bpc_avx512icl:      21.0 ( 8.70x)    21.1 ( 7.53x)

intra_pred_z3_w8_16bpc_avx2:           48.7 (11.31x)    44.7 (10.18x)
intra_pred_z3_w8_16bpc_avx512icl:      30.8 (17.86x)    35.7 (12.76x)

intra_pred_z3_w16_16bpc_avx2:         102.1 (15.49x)    93.5 (14.72x)
intra_pred_z3_w16_16bpc_avx512icl:     59.3 (26.65x)    78.5 (17.53x)

intra_pred_z3_w32_16bpc_avx2:         218.2 (17.05x)   205.2 (15.89x)
intra_pred_z3_w32_16bpc_avx512icl:    116.8 (31.84x)   139.5 (23.38x)

intra_pred_z3_w64_16bpc_avx2:         486.0 (17.52x)   451.2 (16.63x)
intra_pred_z3_w64_16bpc_avx512icl:    322.4 (26.41x)   342.7 (21.89x)