Search results

Update cuda.coop two-phase API to use factory functions

Status: Open.

Enhancement

#7362 In NVIDIA/cccl;

· jrhemstad opened on Jan 26, 2026

Add <code>cub::DeviceTransform</code> benchmarks based on use in pytorch

Status: Open.

#7361 In NVIDIA/cccl;

· bernhardmgruber opened on Jan 26, 2026

Provide an installation path that doesn't install <code>numba-cuda</code>

Status: Open.

#7355 In NVIDIA/cccl;

· shwina opened on Jan 26, 2026

Enable pre-compiled ops in <code>cuda.compute</code>

Status: Open.

#7354 In NVIDIA/cccl;

· shwina opened on Jan 26, 2026

Rewrite iterators to not use <code>numba-cuda</code>

Status: Open.

#7353 In NVIDIA/cccl;

· shwina opened on Jan 26, 2026

Decouple <code>numba-cuda</code> from type system and other internals

Status: Open.

#7352 In NVIDIA/cccl;

· shwina opened on Jan 26, 2026

[FEA]: Provide a way to create an uninitialized array

cubFor all items related to CUBFor all items related to CUB

Status: Open.

Enhancement

#7338 In NVIDIA/cccl;

· miscco opened on Jan 23, 2026

[FEA]: support system ctk install for python package

cuda.computeFor all items related to the cuda.parallel Python moduleFor all items related to the cuda.parallel Python module

Status: Open.

#7328 In NVIDIA/cccl;

· ZzEeKkAa opened on Jan 23, 2026

Add benchmarks for all algorithms (most are there already)

Status: Open.

#7319 In NVIDIA/cccl;

· shwina opened on Jan 21, 2026

Migrate benchmarks from <code>pytest-benchmark</code> to <code>nvbench</code> (https://github.com/NVIDIA/nvbench/)

Status: Open.

#7318 In NVIDIA/cccl;

· shwina opened on Jan 21, 2026

Benchmarking <code>cuda.compute</code>

Status: Open.

#7317 In NVIDIA/cccl;

· shwina opened on Jan 21, 2026

[FEA]: Reenable overflow builtins for nvhpc 26.1

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

Status: Open.

Enhancement

#7303 In NVIDIA/cccl;

· davebayer opened on Jan 21, 2026