All pull requests

Pull requests

• Build and test with CUDA 13.4.3

  improvementnon-breaking
  #3196·jameslamb opened Oct 29, 2026Member·

  ·
  5

• Coalesced small-nkeys kernel for reduce_rows_by_key

  #3195·cakedev0 opened Oct 28, 2026·

  ·
  2

• Use the small-nkeys kernel of reduce_rows_by_key only with few columns

  #3194·cakedev0 opened Oct 28, 2026·

  ·
  4

• Factor shared reduction and linewise kernel launches into host helpers

  #3193·divyegala opened Oct 27, 2026Contributor·

  ·
  1

• Fix make_blobs shuffling and return generated centers

  #3192·NIne-WIngEd opened Oct 27, 2026Contributor·

  ·
  2

  1

• Keep uniform half random values in [0, 1)

  #3190·ViderStar opened Oct 25, 2026·

  ·
  1

  4

• [WIP] Keep randomized SVD headers out of PCA

  #3187·divyegala opened Oct 22, 2026Contributor·

  ·
  1

• Add opt-in compact diagnostic locations

  improvementnon-breaking
  #3184·divyegala opened Oct 21, 2026Contributor·

  ·
  3

• Retry failed Lanczos eigensoves with jacobi

  #3183·JohnZed opened Oct 20, 2026Contributor·

  ·
  1

• Use rapids-github-fetch in ci/check_style.sh

  improvementnon-breaking
  #3181·bdice opened Oct 19, 2026Contributor·

  ·
  1

• Include <cstdint> in raft/core/comms.hpp

  bugnon-breaking
  #3178·ethancedwards8 opened Oct 16, 2026·

  ·
  4

• Optimize sum_rows_by_key() kernel

  improvementnon-breaking
  #3176·huuanhhuyn opened Oct 15, 2026Contributor·

  ·
  25

• Seed pre-commit cache from main builds

  improvementnon-breaking
  #3166·bdice opened Oct 1, 2026Contributor·

  ·
  1

• Use CCCL aliases and stream-ordered pinned and managed resources

  breakingimprovement
  #3159·bdice opened Sep 29, 2026Contributor·

  ·
  1

• Lanczos dense fallback on n <= 3

  breakingbug
  #3122·aamijar opened Aug 20, 2026Member·

  ·
  1

  7

  

• add-batched-linalg for cuvs

  #3110·Intron7 opened Aug 12, 2026Contributor·

  ·
  10

• Fix RAFT_LOG_TRACE_VEC after the rapids-logger migration

  #3105·Wint3rNight opened Aug 6, 2026·

  ·
  1

  4

• chore(ci): use explicit secrets where possible

  improvementnon-breaking
  #3042·gforsyth opened May 27, 2026Contributor·

  ·
  1

  

• Drop thrust::zip_iterator in favor of cuda::zip_iterator

  improvementnon-breaking
  #2964·miscco opened Feb 23, 2026Contributor·

  ·
  2

  

• Add 1xtfloat capability to pairwise_matrix distance computations

  cppimprovementnon-breaking
  #1493·ahendriksen opened May 8, 2023Contributor·

  ·
  11

  

• [FEA] Add sparse LOBPCG solver

  2 - In ProgressCMakecppimprovementnon-breaking
  #1081·lowener opened Dec 8, 2022Contributor·

  ·
  

• Moving low-level iterative solver primitives to raft::solver

  CMakecppimprovementnon-breaking
  #923·cjnolet opened Oct 17, 2022Contributor·

  ·
  1