On Alpha processors, converts the expression on the right-hand
    side of the equal sign to the IEEE T_floating type (double
    precision, length 8 bytes) and deposits the result into the
    specified location.