Issues

• [BUG] cuVS Bench Faiss CAGRA b_ratio reads the wrong JSON key

  bug
  NVIDIA/cuvs#2718 · Filter by author kshitizgupta21 opened on Oct 1, 2026

• [FEA] Move filters out of common.hpp

  feature request
  NVIDIA/cuvs#2714 · Filter by author HowardHuang1 opened on Oct 1, 2026

• [BUG] Lucene: GPU search codecs ignore the filter cache config on read and don't share a cache

  bugLucene
  NVIDIA/cuvs#2705 · Filter by author imotov opened on Sep 30, 2026
  

• [DOC] CAGRA documentation fix for load

  doc
  NVIDIA/cuvs#2695 · Filter by author SuriyaaMM opened on Sep 26, 2026

• Audit cagra.hpp template functions

  improvement
  NVIDIA/cuvs#2691 · Filter by author aamijar opened on Sep 24, 2026
  1

• Remove composite_index

  improvement
  NVIDIA/cuvs#2690 · Filter by author aamijar opened on Sep 24, 2026

• Upgrade faiss to v1.15.1

  faissimprovement
  NVIDIA/cuvs#2689 · Filter by author aamijar opened on Sep 24, 2026

• [BUG] CAGRA normal-hash sizing narrows visited-node count to uint32_t

  NVIDIA/cuvs#2684 · Filter by author shaunakkapur opened on Sep 23, 2026

• [FEA] Out-of-core Smart Merge

  feature request
  NVIDIA/cuvs#2683 · Filter by author cjnolet opened on Sep 23, 2026

• [FEA] Build CAGRA index using PIPNN

  feature request
  NVIDIA/cuvs#2682 · Filter by author cjnolet opened on Sep 23, 2026

• [FEA] Build Vamana index with PIPNN

  feature request
  NVIDIA/cuvs#2681 · Filter by author cjnolet opened on Sep 23, 2026

• Doc gap: no discoverable notice that CAGRA's C++ build API changed argument types in 26.10

  NVIDIA/cuvs#2673 · Filter by author zbrad opened on Sep 22, 2026

• [BUG] cuVS: Refactor CagraIndex Builder with better validation

  bugJava
  NVIDIA/cuvs#2672 · Filter by author imotov opened on Sep 22, 2026
  

• [BUG] cuvs-java: audit exception paths for native resource leaks

  bugJava
  NVIDIA/cuvs#2670 · Filter by author imotov opened on Sep 21, 2026
  

• [Faiss] Large batch out-of-core CAGRA search

  feature request
  NVIDIA/cuvs#2668 · Filter by author cjnolet opened on Sep 21, 2026

• Large batch out-of-core CAGRA search

  feature request
  NVIDIA/cuvs#2667 · Filter by author cjnolet opened on Sep 21, 2026

• Unify NN Descent GNND::build() BBQ and dense versions

  improvement
  NVIDIA/cuvs#2663 · Filter by author aamijar opened on Sep 21, 2026

• [BUG] NNTest: GTEST_SKIP() in compute_1nn() doesn't skip the test, so cuTile-backend cases fail when cuTile is unavailable

  NVIDIA/cuvs#2660 · Filter by author zbrad opened on Sep 18, 2026

• [BUG] kmeans::fit is not run-to-run reproducible; KmeansFitBatchedTestF.Result/4 flakes ~50% on GB10 and should not gate CI

  NVIDIA/cuvs#2657 · Filter by author zbrad opened on Sep 18, 2026

• Disable AnnNNDescentTestUI8_U32 until release/26.10 is complete

  flaky
  NVIDIA/cuvs#2655 · Filter by author aamijar opened on Sep 18, 2026
  

• [DOC] Update published docs in nightly CI

  doc
  NVIDIA/cuvs#2646 · Filter by author cjnolet opened on Sep 17, 2026

• [BUG] CPU fallback for the quantized accelerated-HNSW Lucene codecs throws instead of falling back

  bugLucene
  NVIDIA/cuvs#2639 · Filter by author imotov opened on Sep 17, 2026
  1

• [CI] Run the cuVS Bench Lucene E2E suite with provisioned artifacts

  NVIDIA/cuvs#2635 · Filter by author nvzm123 opened on Sep 16, 2026

• [DOC] Lucene: clarify codec behavior and Java test commands

  NVIDIA/cuvs#2623 · Filter by author nvzm123 opened on Sep 16, 2026

• [BUG] Lucene: close per-query CAGRA device matrices

  NVIDIA/cuvs#2621 · Filter by author nvzm123 opened on Sep 16, 2026