T88: optimized intrinsic

Benchmark                          (size)  Mode  Cnt     Score       Error  Units
ArrayEqualsBench.byteArrayEquals        1  avgt    5      98.554 ?   7.963  ns/op
ArrayEqualsBench.byteArrayEquals       15  avgt    5     108.337 ?   0.491  ns/op
ArrayEqualsBench.byteArrayEquals       32  avgt    5     204.878 ? 264.547  ns/op
ArrayEqualsBench.byteArrayEquals      128  avgt    5     172.961 ?   0.234  ns/op
ArrayEqualsBench.byteArrayEquals      512  avgt    5     684.910 ?   0.709  ns/op
ArrayEqualsBench.byteArrayEquals     1024  avgt    5    1669.154 ? 309.943  ns/op
ArrayEqualsBench.byteArrayEquals  1000000  avgt    5  293951.532 ? 374.712  ns/op


T88: basline intrinsic
Benchmark                         (size)  Mode  Cnt        Score      Error  Units
ArrayEqualsBench.byteArrayEquals       1  avgt    5      107.446 ?   11.349  ns/op
ArrayEqualsBench.byteArrayEquals      15  avgt    5      171.230 ?   67.104  ns/op
ArrayEqualsBench.byteArrayEquals      32  avgt    5      238.655 ?   56.341  ns/op
ArrayEqualsBench.byteArrayEquals     128  avgt    5      301.053 ?    1.136  ns/op
ArrayEqualsBench.byteArrayEquals     512  avgt    5     2095.786 ?  121.570  ns/op
ArrayEqualsBench.byteArrayEquals    1024  avgt    5     2373.010 ? 3795.149  ns/op
ArrayEqualsBench.byteArrayEquals 1000000  avgt    5  1740280.432 ? 4930.242  ns/op