Odd but Error-Free FastTwoSum: More General Conditions for FastTwoSum as an Error-Free Transformation for Faithful Rounding Modes
Sehyeok Park, Jay P. Lim, Santosh Nagarakatte

TL;DR
This paper broadens the applicability of FastTwoSum as an error-free transformation under various faithful rounding modes, including round-to-odd, by establishing more general conditions and tailored floating-point splitting methods.
Contribution
It introduces more general conditions for FastTwoSum as an EFT across all faithful rounding modes and develops a tailored floating-point splitting for round-to-odd modes.
Findings
More general conditions for FastTwoSum as EFT under faithful rounding modes
Guarantees established for round-to-odd modes
A configurable floating-point splitting tailored for round-to-odd modes
Abstract
This paper proposes sufficient, yet more general conditions for applying FastTwoSum as an error-free transformation (EFT) under all faithful rounding modes. Additionally, it also identifies guarantees tailored to round-to-odd for establishing FastTwoSum as an EFT. This paper also describes a floating-point splitting tailored for round-to-odd that is an EFT where the distribution of bits is configurable (i.e., ExtractScalar for round-to-odd). Our sufficient conditions are more general than those previously known in the literature (i.e., it applies to a wider operand domain).
Peer Reviews
No public reviews on file for this paper yet. If you reviewed it on a platform where reviews are public (OpenReview, ICLR, NeurIPS, ICML), you can paste yours below so the community can read it here.
Videos
No videos yet. Explain this paper in a talk, walkthrough, or lecture? Add one.
Taxonomy
TopicsNumerical Methods and Algorithms · Low-power high-performance VLSI design · Digital Filter Design and Implementation
