PDP-11 Floating Point Format

This document, dated October 1, 1970, outlines the standardized floating point format for the PDP-11 series. The format, using sign-magnitude representation with a binary normalized fraction and an excess-128 exponent, was adopted for both hardware implementation in the PDP-11/40 and software packages in the PDP-11/20. The document details the bit allocation for single and double precision representations, noting that the highest-order bit of the fraction is implicitly 1 due to normalization, resulting in 24 bits of precision for single precision and 56 bits for double precision.