saber.yaml

stateful: false
name: SABER
description: Based on Mod-LWR (Module Learning With Rounding).
category: lattice
year:
  candidate: 2017
authors:
  - Jan-Pieter D’Anvers
  - Angshuman Karmakar
  - Sujoy Sinha Roy
  - Frederik Vercauteren
  - Andrea Basso
  - Jose Maria Bermudo Mera
  - Michiel Van Beirendonck
problems:
  - assumption: Mod-LWR
sources:
  - >-
    Round 3 Specification
    (https://www.esat.kuleuven.be/cosic/pqcrypto/saber/files/saberspecround3.pdf)
website: 'https://www.esat.kuleuven.be/cosic/pqcrypto/saber/index.html'
nist round: 3f
last_updated: '2020-11-14'
patents: >-
  Patent US9094189: The patent holder claims that the patent applies to "one or
  more proposals" that were submitted to the NIST PQC standardization process
  [1]. Damien Stehlé and Vadim Lyubashevsky released an analysis followed by the
  claim that this patent does not apply to Kyber and SABER [2] which was
  subsequently discussed by peers [3]. Patents CN108173643 and CN107566121:
  According to the patent holder, SABER falls into their AKCN mechanism for LWR
  schemes. The inventor issued an informal statement in 2022 that they want to
  give up these patents and they do not hold them for economic reasons [4]. More
  patent discussion can be found in [5].
patents sources:
  - https://web.archive.org/web/20220318084318/https://www.cnrsinnovation.com/?lang=en
  - 'https://eprint.iacr.org/2021/1364.pdf'
  - https://groups.google.com/a/list.nist.gov/g/pqc-forum/c/2Xv0mrF9IVo/m/e0oKQhllBwAJ
  - https://groups.google.com/a/list.nist.gov/g/pqc-forum/c/Fm4cDfsx65s/m/aj31YoWWBAAJ
  - https://groups.google.com/a/list.nist.gov/g/pqc-forum/c/6Psr4bFHHgk/m/3rDADGcMAQAJ