NVIDIA/cuvs

View on GitHub

cuVS - a library for vector search and clustering on the GPU

Stars: 854
Forks: 236
Open beginner issues: 0
Indexed issues: 593
Avg merge: 3d 3h
Merged PRs (30d): 62

Dominant language: Cuda
License: Apache-2.0
Last GitHub push: Sep 19, 2026
Latest indexed: Sep 19, 2026
Contributing guide: Contributing guide
Code of conduct: Code of conduct
Beginner labels: No beginner labels indexed

Beginner friendly Everything

Open All

593 issues indexed so far

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

Difficulty 1/5 1-3 hours Newbie friendliness 88/100

NVIDIA/cuvs#2660 · Opened Sep 18, 2026
[BUG] kmeans::fit is not run-to-run reproducible; KmeansFitBatchedTestF.Result/4 flakes ~50% on GB10 and should not gate CI Open

Difficulty 4/5 3-5 days Newbie friendliness 55/100

NVIDIA/cuvs#2657 · Opened Sep 18, 2026
Disable AnnNNDescentTestUI8_U32 until release/26.10 is complete Open

NVIDIA/cuvs#2655 · 1 assignee · Opened Sep 18, 2026
[DOC] Update published docs in nightly CI Open

doc

Difficulty 3/5 1-2 days Newbie friendliness 48/100

NVIDIA/cuvs#2646 · Opened Sep 17, 2026
[BUG] CPU fallback for the quantized accelerated-HNSW Lucene codecs throws instead of falling back Open

bug Lucene

Difficulty 3/5 1-2 days Newbie friendliness 68/100

NVIDIA/cuvs#2639 · 1 comment · Opened Sep 17, 2026
[CI] Run the cuVS Bench Lucene E2E suite with provisioned artifacts Open

Difficulty 5/5 Over a week Newbie friendliness 35/100

NVIDIA/cuvs#2635 · Opened Sep 16, 2026
[DOC] Lucene: clarify codec behavior and Java test commands Open

Difficulty 2/5 1-3 hours Newbie friendliness 82/100

NVIDIA/cuvs#2623 · Opened Sep 16, 2026
[BUG] Lucene: close per-query CAGRA device matrices Open

Difficulty 3/5 1-2 days Newbie friendliness 68/100

NVIDIA/cuvs#2621 · Opened Sep 16, 2026
[FEA] Lucene: provide a thin-JAR adapter for configured PyLucene HNSW Open

Difficulty 4/5 3-5 days Newbie friendliness 68/100

NVIDIA/cuvs#2619 · Opened Sep 16, 2026
[BUG] Lucene: route explicit MULTI_KERNEL queries through per-segment search Open

Difficulty 3/5 1-2 days Newbie friendliness 72/100

NVIDIA/cuvs#2620 · Opened Sep 16, 2026
[BUG] Lucene: preserve released binary-vector indexes while enabling binary storage Open

Difficulty 5/5 Over a week Newbie friendliness 35/100

NVIDIA/cuvs#2617 · Opened Sep 16, 2026
[BUG] Lucene: keep thin-JAR SPI registrations self-owned Open

Difficulty 4/5 3-5 days Newbie friendliness 55/100

NVIDIA/cuvs#2618 · Opened Sep 16, 2026
[IMP] Lucene: initialize configured accelerated codecs once Open

Difficulty 3/5 1-2 days Newbie friendliness 65/100

NVIDIA/cuvs#2616 · Opened Sep 16, 2026
[BUG] Lucene: isolate versioned providers and repair quantized format factories Open

Difficulty 5/5 Over a week Newbie friendliness 42/100

NVIDIA/cuvs#2615 · Opened Sep 16, 2026
[BUG] Lucene: defer vector-format dependency initialization until use Open

Difficulty 4/5 3-5 days Newbie friendliness 45/100

NVIDIA/cuvs#2614 · Opened Sep 16, 2026
[FEA] Remove cuvsCagraCompressionParams and only use cuvsProductQuantizerParams Open

feature request

Difficulty 4/5 3-5 days Newbie friendliness 55/100

NVIDIA/cuvs#2613 · 2 comments · Opened Sep 16, 2026
Iterative CAGRA build temp memory space for graph Open

feature request

Difficulty 4/5 3-5 days Newbie friendliness 30/100

NVIDIA/cuvs#2610 · 1 comment · Opened Sep 15, 2026
[FEA] Reduce duplicate distance computations in Vamana GreedySearch build Open

Difficulty 4/5 3-5 days Newbie friendliness 52/100

NVIDIA/cuvs#2604 · Opened Sep 14, 2026
[CI] CUTILE_SMOKE_TEST GB300 failed Open

bug

Difficulty 3/5 1-2 days Newbie friendliness 55/100

NVIDIA/cuvs#2602 · Opened Sep 14, 2026
[DOC] RAPIDS -> nvidia docs migration TODOs Open

doc

Difficulty 4/5 3-5 days Newbie friendliness 50/100

NVIDIA/cuvs#2598 · Opened Sep 11, 2026
WIP: Optimize CAGRA and/or CAGRA_HNSW FBIN bulk indexing Open

Difficulty 5/5 Over a week Newbie friendliness 25/100

NVIDIA/cuvs#2592 · Opened Sep 11, 2026
Reuse reconstruction functions from PQ api in Iterative CAGRA-Q Open

Difficulty 4/5 3-5 days Newbie friendliness 35/100

NVIDIA/cuvs#2591 · Opened Sep 10, 2026
Float codebook vectors support for PQ dataset and Iterative CAGRA-Q Open

Difficulty 5/5 Over a week Newbie friendliness 35/100

NVIDIA/cuvs#2590 · Opened Sep 10, 2026
[CI] NEIGHBORS_MG_TEST flaky Open

bug

Difficulty 4/5 3-5 days Newbie friendliness 35/100

NVIDIA/cuvs#2588 · Opened Sep 10, 2026
[BUG] cuVS 26.08 CAGRA incompatibility with FAISS v1.14.3 Open

bug

Difficulty 4/5 3-5 days Newbie friendliness 52/100

NVIDIA/cuvs#2587 · 6 comments · Opened Sep 10, 2026
[FEA] Combine KMeans and ANN Batch Loaders Open

feature request

NVIDIA/cuvs#2581 · 1 assignee · Opened Sep 10, 2026
Reduce combinatorial GTest coverage using explicit size buckets and metric sharding Open

Difficulty 5/5 Over a week Newbie friendliness 35/100

NVIDIA/cuvs#2578 · Opened Sep 9, 2026
[FEA] Rename cagraCompressionParams to cuvsPqParams Open

feature request

Difficulty 5/5 Over a week Newbie friendliness 35/100

NVIDIA/cuvs#2576 · Opened Sep 9, 2026
[FEA] Iterative CAGRA-Q in ANN_BENCH Open

feature request

Difficulty 4/5 3-5 days Newbie friendliness 35/100

NVIDIA/cuvs#2575 · Opened Sep 9, 2026
Padded dataset can affect IVF-PQ encoding Open

Difficulty 4/5 3-5 days Newbie friendliness 38/100

NVIDIA/cuvs#2569 · Opened Sep 9, 2026
PQ preprocessor does not natively support padded datasets Open

feature request

Difficulty 3/5 1-2 days Newbie friendliness 58/100

NVIDIA/cuvs#2567 · Opened Sep 9, 2026
Possible memory leak: Tensor.Close returns early on the CUDA path and skips two frees Open

Difficulty 2/5 1-3 hours Newbie friendliness 84/100

NVIDIA/cuvs#2558 · Opened Sep 7, 2026
[CI] TestCuVSAcceleratedHNSWGaps.testVectorSearchWithAlternatingDocuments is flaky Open

ci Lucene

Difficulty 2/5 1-3 hours Newbie friendliness 72/100

NVIDIA/cuvs#2551 · Opened Sep 3, 2026
[CI] SCANN Test flaky Open

bug

Difficulty 3/5 1-2 days Newbie friendliness 52/100

NVIDIA/cuvs#2549 · Opened Sep 3, 2026
[BUG] ivf_rabitq needs seralize & deserilize steps before the index is searchable Open

bug

Difficulty 3/5 1-2 days Newbie friendliness 68/100

NVIDIA/cuvs#2537 · Opened Aug 31, 2026
[BUG] CAGRA: calc_hashmap_params() never terminates for large itopk_size (hash_bitlen limit is checked after the sizing loop) Open

Difficulty 3/5 1-2 days Newbie friendliness 74/100

NVIDIA/cuvs#2523 · Opened Aug 28, 2026
Lucene: replace ExecutorService with StructuredTaskScope Open

bug Lucene

NVIDIA/cuvs#2518 · 1 assignee · Opened Aug 28, 2026
Upgrade to JDK 25 as the minimal version for using cuvs-java and cuvs-lucene Open

improvement Java

Difficulty 3/5 1-2 days Newbie friendliness 55/100

NVIDIA/cuvs#2517 · Opened Aug 28, 2026
[Tracker] Improving cuvs-lucene release robustness Open

Difficulty 5/5 Over a week Newbie friendliness 20/100

NVIDIA/cuvs#2512 · Opened Aug 27, 2026
Use half2 and float2 types/operators for Vamana build L2 metrics Open

Difficulty 4/5 3-5 days Newbie friendliness 45/100

NVIDIA/cuvs#2510 · Opened Aug 26, 2026
[FEA] Index build benchmark in big-ann-benchmarks Open

feature request

Difficulty 4/5 3-5 days Newbie friendliness 45/100

NVIDIA/cuvs#2508 · Opened Aug 26, 2026
[BUG] Downloading and decompressing cuvs_conda_cpp_libcuvs_x86_64_cu12 occasionally takes a very long time Open

bug Build

Difficulty 3/5 1-2 days Newbie friendliness 55/100

NVIDIA/cuvs#2500 · Opened Aug 22, 2026
Replace Vamana build scatter kernel with RAFT primitives Open

Difficulty 4/5 3-5 days Newbie friendliness 48/100

NVIDIA/cuvs#2493 · Opened Aug 19, 2026
Accept PQ datasets in `extend()` CAGRA Open

NVIDIA/cuvs#2492 · 1 assignee · Opened Aug 19, 2026
[BUG] cagra::search on vpq_f16_index triggers OOB for every search_algo (26.10 nightly) Open

bug

Difficulty 4/5 3-5 days Newbie friendliness 48/100

NVIDIA/cuvs#2489 · 1 comment · Opened Aug 19, 2026
[BUG] cuvsDatasetMakePadded rejects an already-aligned device tensor Open

bug

Difficulty 4/5 3-5 days Newbie friendliness 48/100

NVIDIA/cuvs#2482 · 1 reaction · Opened Aug 18, 2026
[BUG] Host-data connect_knn_graph passes a non-symmetric CSR to RAFT's MST Open

bug

Difficulty 3/5 1-2 days Newbie friendliness 74/100

NVIDIA/cuvs#2477 · Opened Aug 18, 2026
Pass dataset views by value instead of `const&` Open

NVIDIA/cuvs#2445 · 1 assignee · Opened Aug 14, 2026
[FEA][C++][C][Language Wrappers] Rename vpq to pq for datasets, factories, and params Open

feature request

NVIDIA/cuvs#2440 · 1 assignee · Opened Aug 12, 2026
Tiered index converts to padded on every search call Open

NVIDIA/cuvs#2434 · 1 assignee · Opened Aug 12, 2026

NVIDIA/cuvs

Get new issues in your inbox