* `f16v2add $aDst0, $aSrc0, $aSrc1`
* `f16v2add $aDst0, $aSrc0:BL, $aSrc1`
* `f16v2add $aDst0, $aSrc0:BU, $aSrc1`

*Half-precision* floating-point vector add on two register source
values.