* `f32v4tof16 $aDst0:Dst0+1, $aSrc0:Src0+3`

*Single-precision* floating-point 4-element vector to *f16* 4-element
vector conversion.