CADO-NFS, an implementation of the number field sieve for integer factorization, registered at APP under number IDDN.FR.001.150006.000.S.P.2010.000.10000.
GF2X [gitlab], a library for multiplication in GF(2)[x] (with Richard P. Brent, Pierrick Gaudry and Emmanuel Thomé). Also contains code to search for irreducible trinomials over GF(2) [replaces and superseedes the irred-ntl code].
mul_fft-6.1.2.tgz, a GMP-based implementation of Schönhage-Strassen's large integer multiplication algorithm (with Pierrick Gaudry and Alexander Kruppa). This code is up to26% faster than the one distributed within GMP 6.1.2. It was used by Andrew Sutherland to set a new record for genus 1 point-counting over a prime field. See also the mirror on Gaudry's web page. A fixed-length FFT code is also available for length 26 [see corresponding mail].
GMP-ECM: an implementation of Lenstra's Elliptic Curve Method to factor integers. See also the ECMNET page.
<hjacobi.c> implements an O(M(n) log n) algorithm for the Jacobi symbol (joint work with Richard Brent).
<double%5Ffac%5Fui.c> implements the double factorial (see this and that).
WCLR implements the algorithm described in the articles Worst Cases and Lattice Reduction (with Vincent Lefèvre). It is now superseeded by Bacsel (with Guillaume Hanrot, Vincent Lefèvre, and Damien Stehlé) [gitlab]
DPE, a library for floating-point arithmetic with large exponents (double-precision + exponent), developed with Patrick Pélissier. [Software Heritage archive, cf thispaper]
Epelle: a generic spell-checking mechanism, including a french dictionary (but it can be used for any language, as soon as you have a dictionary).
<intersectplot.mpl>, a Maple program to plot the intersection of two 3D surfaces given by implicit equations (with Sylvain Petitjean)
MPCHECK 1.1.0, with Nathalie Revol and Patrick Pélissier, a program to check mathematical function libraries (correct rounding, monotonicity, symmetry, output range). [results] Warning: version 1.1.0 only works for the double type (53 bits). [gitlab]
<shift2.c>, a implementation in GNU MP of an asymptotically fast algorithm for Taylor shift. Werner Krandick reports a speedup factor of 319 for degree 9,000 and coefficients of 9,000 bits, over von zur Gathen's and Gerhard's convolution method. This work was inspired by High-Performance Implementations of the Descartes Method, by Jeremy R. Johnson, Werner Krandick, Kevin M. Lynch, David G. Richardson, and Anatole D. Ruslanov.