A Novel Data Format for Approximate Arithmetic Computing

Approximate computing has become one of the most popular computing paradigms in the era of the Internet of things and big data. It takes advantages of the error-tolerable feature of many applications, such as machine learning and image/signal processing, to reduce the resource required to deliver certain level of computation quality. In this paper, we propose an approximate integer format (AIF) and its associated arithmetic operations for energy minimization with controllable computation accuracy. In AIF, operands are segmented at run time such that the computation is performed only on part of operands by computing units (such as adders and multipliers) of smaller bit-width. The proposed AIF can be used for any arithmetic operation and can be extended to fixed point numbers. It can also be incorporated into higher level design such as architectural and programming language to give user the control of approximate computing. Experimental results show that our AIF based approximation computing approach can achieve high accuracy, incurs very little additional overhead, and save considerable energy.