# f16v4sum $aDst0:Dst0+1 $aSrc0:Src0+1

*Half-precision* vector 2x2 lateral summation