Issues

• [BUG]: libcu++ NVRTC header parsing cost and PCH footprint

  needs triage
  Bug

  NVIDIA/cccl#12057 · Filter by author lamarrr opened on Oct 17, 2026
  2

  

• [FEA]: Make __resizable_buffer public in some capacity

  NVIDIA/cccl#12051 · Filter by author KyleFromNVIDIA opened on Oct 16, 2026

• [FEA]: a typed pipeline for cuda.coop on Numba-CUDA-MLIR

  cuda.coopneeds triage
  NVIDIA/cccl#12022 · Filter by author diptorupd opened on Oct 15, 2026
  1

  

• [BUG]: Using ProxyArray with cuda.compute.make_merge_sort fails to compile

  Bug

  NVIDIA/cccl#12021 · Filter by author oleksandr-pavlyk opened on Oct 15, 2026
  1

• [CUB] Track warp reduction optimizations

  NVIDIA/cccl#11999 · Filter by author fbusato opened on Oct 13, 2026
  2

• Investigate implementing some of c2h helpers' machine for cuda::buffer support to the CUDA buffer or memory resources

  NVIDIA/cccl#11990 · Filter by author oleksandr-pavlyk opened on Oct 12, 2026

• [BUG]: Incorrect test asserts in libcudacxx/test/libcudacxx/std - they constant evaluate on Clang

  needs triage
  Bug

  NVIDIA/cccl#11989 · Filter by author seanbaxter opened on Oct 12, 2026
  2

  

• Re-tune DeviceBatchedTopK

  NVIDIA/cccl#11988 · Filter by author elstehle opened on Oct 12, 2026

• DeviceBatchedTopK Perform twiddle-in just "once" on read and identify solution for the re-streamed items

  NVIDIA/cccl#11987 · Filter by author elstehle opened on Oct 12, 2026

• Improve +/-0.0 twiddling logic (to recoup performance from re-mapping -0.0 to +0.0)

  NVIDIA/cccl#11971 · Filter by author elstehle opened on Oct 9, 2026
  

• Fix no-throw CUDA Driver APIs

  NVIDIA/cccl#11970 · Filter by author pciolkosz opened on Oct 9, 2026
  

• PR benchmark CI should support Thrust and libcu++ benchmarks

  NVIDIA/cccl#11965 · Filter by author bernhardmgruber opened on Oct 9, 2026

• Recoup performance in DeviceBatchedTopK after +/-0.0 twiddling fix

  NVIDIA/cccl#11960 · Filter by author elstehle opened on Oct 9, 2026
  

• Thread cub::BlockMerge into cub::BlockMergeSort()

  cub
  Enhancement

  NVIDIA/cccl#11949 · Filter by author Jacobfaib opened on Oct 8, 2026
  

• Thread the new cub::BlockMerge() into cub::DeviceMergeSort()

  cub
  Enhancement

  NVIDIA/cccl#11948 · Filter by author Jacobfaib opened on Oct 8, 2026
  

• Add implementation of NCCL device MGMN sort

  cudaxMGMN
  Enhancement

  NVIDIA/cccl#11947 · Filter by author Jacobfaib opened on Oct 8, 2026
  

• Add implementation of cub::DeviceMerge()

  cub
  Enhancement

  NVIDIA/cccl#11946 · Filter by author Jacobfaib opened on Oct 8, 2026
  

• Add implementation of cub::BlockMerge()

  cub
  Enhancement

  NVIDIA/cccl#11945 · Filter by author Jacobfaib opened on Oct 8, 2026
  

• [BUG]: DeviceHistogram::HistogramEven bins samples just below upper_level as num_bins: lost counts, and an out-of-bounds write for unsigned 8-bit samples

  needs triage
  NVIDIA/cccl#11937 · Filter by author builtin-pb opened on Oct 8, 2026
  2

  

• Add CUDA Forward Compatibility package to our CI runners to allow testing new CUDA features

  NVIDIA/cccl#11935 · Filter by author pciolkosz opened on Oct 8, 2026

• Add 64-bit counter and offset coverage to histogram benchmarks

  NVIDIA/cccl#11927 · Filter by author gonidelis opened on Oct 7, 2026
  1

  

• Guard histogram benchmarks against offset and row-stride overflow.

  NVIDIA/cccl#11925 · Filter by author gonidelis opened on Oct 7, 2026
  

• Use fast integer division in HistogramEven

  NVIDIA/cccl#11924 · Filter by author gonidelis opened on Oct 7, 2026
  

• Remove unused CCCL headers

  NVIDIA/cccl#11921 · Filter by author fbusato opened on Oct 7, 2026
  

• [FEA]: PoC fancy iterator unpacking facility

  NVIDIA/cccl#11917 · Filter by author gevtushenko opened on Oct 7, 2026
  3