DeepAI
Log In Sign Up

RLIBM-32: High Performance Correctly Rounded Math Libraries for 32-bit Floating Point Representations

04/08/2021
by   Jay P. Lim, et al.
0

This paper proposes a set of techniques to develop correctly rounded math libraries for 32-bit float and posit types. It enhances our RLibm approach that frames the problem of generating correctly rounded libraries as a linear programming problem in the context of 16-bit types to scale to 32-bit types. Specifically, this paper proposes new algorithms to (1) generate polynomials that produce correctly rounded outputs for all inputs using counterexample guided polynomial generation, (2) generate efficient piecewise polynomials with bit-pattern based domain splitting, and (3) deduce the amount of freedom available to produce correct results when range reduction involves multiple elementary functions. The resultant math library for the 32-bit float type is faster than state-of-the-art math libraries while producing the correct output for all inputs. We have also developed a set of correctly rounded elementary functions for 32-bit posits.

READ FULL TEXT
07/09/2020

A Novel Approach to Generate Correctly Rounded Math Libraries for New Floating Point Representations

Given the importance of floating-point (FP) performance in numerous doma...
11/25/2021

RLIBM-PROG: Progressive Polynomial Approximations for Fast Correctly Rounded Math Libraries

This paper presents a novel method for generating a single polynomial ap...
01/13/2021

FBGEMM: Enabling High-Performance Low-Precision Deep Learning Inference

Deep learning models typically use single-precision (FP32) floating poin...
06/17/2021

Fast evaluation of some p-adic transcendental functions

We design algorithms for computing values of many p-adic elementary and ...
03/09/2021

Fast and Efficient Bit-Level Precision Tuning

In this article, we introduce a new technique for precision tuning. This...