NVIDIA

NVIDIA/cuCollections

661 stars 117 forks 661 watchers Cuda Apache License 2.0
cpp cpp17 cuda datastructures gpu hashmap hashset hashtable
View on GitHub
6 Open Issues Need Help Last updated: Aug 20, 2026

Open Issues Need Help

View All on GitHub
[BUG]: Fix issues with hash table constructors and assignment operators. 3 days ago
type: bug good first issue
NVIDIA/cuCollections
661

Cuda
#cpp#cpp17#cuda#datastructures#gpu#hashmap#hashset#hashtable
Relax key/payload/slot type constraints to enable 1B and 2B types 3 months ago
good first issue P2: Nice to have type: improvement
NVIDIA/cuCollections
661

Cuda
#cpp#cpp17#cuda#datastructures#gpu#hashmap#hashset#hashtable
[ENHANCEMENT]: Enable PDL for all cuco kernels 7 months ago
good first issue helps: rapids topic: performance P1: Should have type: improvement
NVIDIA/cuCollections
661

Cuda
#cpp#cpp17#cuda#datastructures#gpu#hashmap#hashset#hashtable
[BUG]: Investigate the mixed-join performance regression caused by register spilling 10 months ago
type: bug help wanted helps: rapids topic: performance
NVIDIA/cuCollections
661

Cuda
#cpp#cpp17#cuda#datastructures#gpu#hashmap#hashset#hashtable
[ENHANCEMENT]: Bypass key equality check for multiset and multimap insertions 11 months ago
type: feature request good first issue topic: static_multimap topic: static_multiset
NVIDIA/cuCollections
661

Cuda
#cpp#cpp17#cuda#datastructures#gpu#hashmap#hashset#hashtable
[ENHANCEMENT]: Add an example for performing a heterogeneous lookup 11 months ago
type: feature request good first issue P1: Should have
NVIDIA/cuCollections
661

Cuda
#cpp#cpp17#cuda#datastructures#gpu#hashmap#hashset#hashtable