Balancing Inexactness in Mixed Precision Matrix Computations

Erin Claire Carson · 2026

Read Balancing Inexactness in Mixed Precision Matrix Computations by Erin Claire Carson (2026) in the Tower of Babel, a free online library of public-domain works preserved in clean, searchable text. No sign-up. The full text is served as a plain-text file at a predictable URL, where text is present: https://antarctic-labs.com/tower-of-babel/arxiv-2602-04348.txt.

Description

Erin Claire Carson argues that sloppy arithmetic is fine when the problem is already sloppy. Her 2026 paper looks at mixed precision computing, where parts of a calculation run at different accuracies on modern hardware. The question is where low precision is safe.

When grid, sensor, or approximation errors already dominate, rounding error from low precision may cost nothing measurable. Carson studies how these error sources interact in matrix computations. Balancing them on purpose lets scientists choose precisions that speed computing without visible accuracy loss. Worked examples from matrix algebra show the promise, timely as hardware more and more rewards mixed precision.

Catalog record

CREATORErin Claire Carson
YEAR2026
COLLECTIONPAPERS
FORMATTXT
FILE SIZE29 KB
VERSION1.0
LICENSEhttp://creativecommons.org/licenses/by/4.0/
RIGHTSCC BY
SOURCEarXiv

Source: https://arxiv.org/abs/2602-04348

Browse the full library · About the Tower of Babel