* `f16v2absmax $aDst0, $aSrc1, $aSrc0`

*Half-precision* floating-point vector max of absolute values