The assembler generates only IEEE floating-point numbers. The older floiating point formats are not supported.