Search results

[cccl.c] Implement reduce_into using host JIT

Status: Open.

#8696 In NVIDIA/cccl;

· shwina opened on Apr 27, 2026

B200 tunings for warpspeed scan caused regression for I8

Status: Open.

#8688 In NVIDIA/cccl;

· bernhardmgruber opened on Apr 26, 2026

Warpspeed scan is not clearly better than lookback scan on RTX PRO 6000

Status: Open.

#8682 In NVIDIA/cccl;

· bernhardmgruber opened on Apr 24, 2026

[FEA]: Provide way API for forcing mapping to produce 64-bit wide ranks/counts

cudaxFeature intended for the cudax experimental libraryFeature intended for the cudax experimental library

Status: Open.

Enhancement

#8660 In NVIDIA/cccl;

· davebayer opened on Apr 24, 2026

Add a unit test that cudaMallocManaged is available and working

Status: Open.

#8649 In NVIDIA/cccl;

· bernhardmgruber opened on Apr 23, 2026

Reenable thrust OpenMP in clang-tidy

Status: Open.

#8647 In NVIDIA/cccl;

· davebayer opened on Apr 23, 2026

[BUG]: cuda::transform_iterator silently downgrades iterator_category → 2.5–3.3× slowdowns for thrust algorithms that dispatch on it

needs triageIssues that require the team's attentionIssues that require the team's attention

Status: Open.

Bug

#8640 In NVIDIA/cccl;

· ttnghia opened on Apr 22, 2026

[BUG] Hierarchy count and rank queries always return size_t

cudaxFeature intended for the cudax experimental libraryFeature intended for the cudax experimental library

libcu++For all items related to libcu++For all items related to libcu++

Status: Open.

Bug

#8637 In NVIDIA/cccl;

· davebayer opened on Apr 22, 2026

cub::DeviceSelect::UniqueByKey might be ambiguous

Status: Open.

#8633 In NVIDIA/cccl;

· gonidelis opened on Apr 22, 2026

[FEA] Use std::scoped_lock instead of std::lock_guard

Status: Open.

Enhancement

#8632 In NVIDIA/cccl;

· davebayer opened on Apr 22, 2026

[INFRA]: Update cccl.c devcontainers to 26.06

infraShared CMake, github, etc infrastructureShared CMake, github, etc infrastructure

Status: Open.

#8620 In NVIDIA/cccl;

· davebayer opened on Apr 22, 2026

Look into cuDF bitmask kernels that would benefit from hierarchical parallelism

Status: Open.

#8598 In NVIDIA/cccl;

· NaderAlAwar opened on Apr 21, 2026