IEEE Standard for Floating-Point Arithmetic

By D. Zuras, M. Cowlishaw, A. Aiken, M. Applegate, D. Bailey, S. Bass, D. Bhandarkar, M. Bhat, David S. Bindel, S. Boldo and 82 more. 2008

2,319 citations recorded by the source corpus.

Subfields: Computational Theory and Mathematics, Electrical and Electronic Engineering, Signal Processing

View Paper: https://www.semanticscholar.org/paper/d2e4fb59b2bf04a0b324db22ed6caf97ffd8f56a

How researchers see this paper

The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.

In Computational Theory and Mathematics

In Electrical and Electronic Engineering

In Signal Processing

Citations in the corpus

Pruning and Quantization for Deep Neural Network Acceleration: A Survey (2021, background): “2) Mini float: bit-width reduced floating-point numbers, which follows the IEEE-754 standard [209].”
Beating Floating Point at its Own Game: Posit Arithmetic (2017, background): “The most recent version (2008) of the IEEE 754 standard [7] includes the fused multiply-add in its requirements.”
Cascade Chaotic System With Applications (2015, methodology): “…to truncate the binary bit stream m from the bit location k 1 to location ( k 1 + 7 ) . x i and y i are 52-bit binary streams converted by outputs x i and y i with the IEEE 754 standard [25]. k 1 and k 2 are two integers deﬁned by k 1 = ( x i × 10 10 mod 6 ) + 39 k 2 = ( y i × 10 10 mod 6 ) + 39 .”
Towards a Standard for Identifying and Managing Bias in Artificial Intelligence (2022, methodology): “Furthermore, to fit these parameters into computer memory, it is often necessary to use half-precision floating-point numbers [161], introducing rounding error which may undermine stability in the numerical methods [162].”
Chain of Thought Empowers Transformers to Solve Inherently Serial Problems (2024, methodology): “Our notion of floating-point number simplifies the IEEE 754 Standard for Floating-point Arith-metic (IEEE, 2008) by removing ∞ and −∞ .”
On significance of the least significant bits for differential privacy (2012, methodology): “A twist on the method above uses a single draw from the uniform distribution on (0 , 1) to generate the sign and the absolute value of the Laplacian variable: All diﬀerentially private systems considered by this paper use one of the two methods (1) and (2) to sample a Laplacian 1 .”
Precimonious: Tuning assistant for floating-point precision (2013, background): “Floating-point arithmetic [15, 26] is used in applications from a wide variety of domains such as high-performance computing, graphics and ﬁnance.”