CUDA Core Compute Libraries (CCCL) Files

Full Changelog: https://github.com/NVIDIA/cccl/compare/v3.3.0...v3.3.0

What's Changed

📚 Libcudacxx

• [libcudacxx] Fix a typo in the documentation by @caugonnet in https://github.com/NVIDIA/cccl/pull/7330
• Add a test for <nv/target> to validate old dialect support. by @wmaxey in https://github.com/NVIDIA/cccl/pull/7241

🔄 Other Changes

• Implement cudax::cufile by @davebayer in https://github.com/NVIDIA/cccl/pull/6122
• Update linear_congruential_generator with constexpr, tests and a fast discard by @RAMitchell in https://github.com/NVIDIA/cccl/pull/6402
• Replace _CCCL_HAS_CUDA_COMPILER() with _CCCL_CUDA_COMPILATION() by @davebayer in https://github.com/NVIDIA/cccl/pull/6399
• Remove unnecessary casts in complex multiplication/division by @davebayer in https://github.com/NVIDIA/cccl/pull/6670
• Add benchmark batch script by @bernhardmgruber in https://github.com/NVIDIA/cccl/pull/6661
• Improvements and testing for inspect_changes CI functionality. by @alliepiper in https://github.com/NVIDIA/cccl/pull/6535
• Improve clarity of CCCL assert macro documentation by @jrhemstad in https://github.com/NVIDIA/cccl/pull/6675
• Fix oversubscription issue with lit precompile, label hack by @alliepiper in https://github.com/NVIDIA/cccl/pull/6554
• Make missing sccache nonfatal. by @alliepiper in https://github.com/NVIDIA/cccl/pull/6582
• Address pending comments for make_tma_descriptor by @fbusato in https://github.com/NVIDIA/cccl/pull/6662
• Add nvhpc 25.9. by @alliepiper in https://github.com/NVIDIA/cccl/pull/6003
• Test building for all arches. by @alliepiper in https://github.com/NVIDIA/cccl/pull/6113
• Add nvbench_helper tests to CI. by @alliepiper in https://github.com/NVIDIA/cccl/pull/6679
• Add more targets to pytorch build. by @alliepiper in https://github.com/NVIDIA/cccl/pull/6685
• Add host std lib version detection by @davebayer in https://github.com/NVIDIA/cccl/pull/6678
• Improve CUB benchmark docs by @bernhardmgruber in https://github.com/NVIDIA/cccl/pull/6640
• Use if consteval in libcu++ by @davebayer in https://github.com/NVIDIA/cccl/pull/6424
• Update docs for _CCCL_IF_CONSTEVAL by @davebayer in https://github.com/NVIDIA/cccl/pull/6692
• Fixes issue with select close to int_max by @elstehle in https://github.com/NVIDIA/cccl/pull/6641
• Update libcudacxx C++ dialect handling. by @alliepiper in https://github.com/NVIDIA/cccl/pull/6693
• Simplifies env usage in DeviceTopK tests by @elstehle in https://github.com/NVIDIA/cccl/pull/6680
• Switch to S3 preprocessor cache by @alliepiper in https://github.com/NVIDIA/cccl/pull/6561
• fix omp scan bug by @charan-003 in https://github.com/NVIDIA/cccl/pull/6560
• Refactor out variant from transform tunings by @bernhardmgruber in https://github.com/NVIDIA/cccl/pull/6669
• [libcu++] Waive hierarchy constexpr testing on GCC8 by @pciolkosz in https://github.com/NVIDIA/cccl/pull/6707
• Use wrapper with void* argument types for iterator advance/dereference signature by @shwina in https://github.com/NVIDIA/cccl/pull/6634
• Restore libcudacxx dialect presets. by @alliepiper in https://github.com/NVIDIA/cccl/pull/6705
• Refactor error handling in radix sort dispatch by @bernhardmgruber in https://github.com/NVIDIA/cccl/pull/6681
• Remove special dialect handling from cudax build system. by @alliepiper in https://github.com/NVIDIA/cccl/pull/6702
• Segmented scan followup by @oleksandr-pavlyk in https://github.com/NVIDIA/cccl/pull/6706
• Fix electing leader from any group in cuda::memcpy_async by @bernhardmgruber in https://github.com/NVIDIA/cccl/pull/6710
• Avoid scaling twice in ReduceNondeterministicPolicy by @bernhardmgruber in https://github.com/NVIDIA/cccl/pull/6711
• Remove special handling of C++ dialect in CUB's build system by @alliepiper in https://github.com/NVIDIA/cccl/pull/6713
• [libcu++] Use resource test fixture members through this by @pciolkosz in https://github.com/NVIDIA/cccl/pull/6717
• Improves top-k examples to illustrate stream usage by @elstehle in https://github.com/NVIDIA/cccl/pull/6723
• Tweak sol.py a bit by @bernhardmgruber in https://github.com/NVIDIA/cccl/pull/6721
• Implement PCG64 as extension by @RAMitchell in https://github.com/NVIDIA/cccl/pull/6292
• Use PDL in cub::DeviceScan by @bernhardmgruber in https://github.com/NVIDIA/cccl/pull/6639
• Fix header in libcudacxx test by @alliepiper in https://github.com/NVIDIA/cccl/pull/6726
• Remove dead code. by @alliepiper in https://github.com/NVIDIA/cccl/pull/6725
• Add deps on thrust/cub to libcudacxx. by @alliepiper in https://github.com/NVIDIA/cccl/pull/6694
• Remove special handling for dialect in Thrust's build system. by @alliepiper in https://github.com/NVIDIA/cccl/pull/6722
• [libcu++] Automatically bump up the release threshold of default mempools by @pciolkosz in https://github.com/NVIDIA/cccl/pull/6718
• Backport cuda::std::reference_wrapper C++20 features by @davebayer in https://github.com/NVIDIA/cccl/pull/6709
• Relax error tolerance for deterministic_device_reduce (RFA) test by @srinivasyadav18 in https://github.com/NVIDIA/cccl/pull/6720
• [DOC] Add temp_storage_bytes usage guide by @Aminsed in https://github.com/NVIDIA/cccl/pull/6208
• Improve charconv test compile times by @davebayer in https://github.com/NVIDIA/cccl/pull/6687
• Move source location builtins directly to <cuda/std/source_location> by @davebayer in https://github.com/NVIDIA/cccl/pull/6738
• Small improvements for cuda::ipow by @davebayer in https://github.com/NVIDIA/cccl/pull/6736
• Add support for clang's alignment builtins by @davebayer in https://github.com/NVIDIA/cccl/pull/6741
• Disable test that is failing in multiple configurations by @miscco in https://github.com/NVIDIA/cccl/pull/6745
• Implement std::normal_distribution by @RAMitchell in https://github.com/NVIDIA/cccl/pull/6585
• Update cuda::std::span concepts by @davebayer in https://github.com/NVIDIA/cccl/pull/6744
• Improve bit builtins support by @davebayer in https://github.com/NVIDIA/cccl/pull/6737
• Implement ranges::drop_view by @miscco in https://github.com/NVIDIA/cccl/pull/5049
• Improve fp decompose by @davebayer in https://github.com/NVIDIA/cccl/pull/6749
• Enable caching of advance/dereference methods for Zipiterator and PermutationIterator by @shwina in https://github.com/NVIDIA/cccl/pull/6753
• implement indeterminate_domain from P3826R2 by @ericniebler in https://github.com/NVIDIA/cccl/pull/6628
• Fix cuda::std::reference_wrapper noexcept test with gcc-8 by @davebayer in https://github.com/NVIDIA/cccl/pull/6757
• cuda.compute: In TransformIterator, use type annotations (if available) to determine the output type of user-provided op by @shwina in https://github.com/NVIDIA/cccl/pull/6760
• cuda.compute: Fixes and improvements to function caching by @shwina in https://github.com/NVIDIA/cccl/pull/6758
• Fix __throw_cuda_error availability with nvrtc by @davebayer in https://github.com/NVIDIA/cccl/pull/6759
• Implement ranges::find_if and ranges::find_if_not by @miscco in https://github.com/NVIDIA/cccl/pull/6752
• Fix radix_sort tuning namespace by @bernhardmgruber in https://github.com/NVIDIA/cccl/pull/6755
• [libcu++] Add sm_62 arch traits by @pciolkosz in https://github.com/NVIDIA/cccl/pull/6772
• fix(readme): Update broken Godbolt example link by @miyanyan in https://github.com/NVIDIA/cccl/pull/6773
• Implement CUDA backend for parallel cuda::std::for_each by @miscco in https://github.com/NVIDIA/cccl/pull/5610
• Ensure that we properly warn about device lambdas that need to query the return type by @miscco in https://github.com/NVIDIA/cccl/pull/6765
• Add missing test for thrust::reduce_into by @Pansysk75 in https://github.com/NVIDIA/cccl/pull/6572
• cuda.compute: Add select algorithm based on three_way_partition by @shwina in https://github.com/NVIDIA/cccl/pull/6766
• Add queries for CUB ptx version as arch_id by @bernhardmgruber in https://github.com/NVIDIA/cccl/pull/6776
• Add operator<< to some CUB enums by @bernhardmgruber in https://github.com/NVIDIA/cccl/pull/6774
• cuda.compute: Fix caching of functions that call other functions by @shwina in https://github.com/NVIDIA/cccl/pull/6770
• Implement std::exponential_distribution by @RAMitchell in https://github.com/NVIDIA/cccl/pull/6750
• Fix issue with libcudacxx header tests. by @alliepiper in https://github.com/NVIDIA/cccl/pull/6785
• Add a type and operation enum to CUB by @bernhardmgruber in https://github.com/NVIDIA/cccl/pull/6780
• Use conventional order of _CCCL_API friend consistently by @miscco in https://github.com/NVIDIA/cccl/pull/6781
• Implement std::binomial_distribution by @RAMitchell in https://github.com/NVIDIA/cccl/pull/6747
• Fixes i32 overflow for benchmark data generation of more than INT_MAX number of items by @elstehle in https://github.com/NVIDIA/cccl/pull/6809
• Temporarily add upper bound to numba-cuda dependency by @shwina in https://github.com/NVIDIA/cccl/pull/6815
• Make cuda capabilities part of cccl config by @davebayer in https://github.com/NVIDIA/cccl/pull/6806
• Update std::uniform_real_distribution by @RAMitchell in https://github.com/NVIDIA/cccl/pull/6798
• [cub] Implement cub::MaxPotentialDynamicSmemBytes by @davebayer in https://github.com/NVIDIA/cccl/pull/6818
• libcudacxx: streamline simple trait aliases by @Aminsed in https://github.com/NVIDIA/cccl/pull/6740
• Fix a typo in compute.rst by @shwina in https://github.com/NVIDIA/cccl/pull/6826
• Improve our WarpReduce implementation by @miscco in https://github.com/NVIDIA/cccl/pull/6814
• Implement cuda::sincos by @davebayer in https://github.com/NVIDIA/cccl/pull/6742
• Replace inline ptx with intrinsics by @davebayer in https://github.com/NVIDIA/cccl/pull/6810
• [cudax->libcu++] Move buffer type from cudax to libcu++ by @pciolkosz in https://github.com/NVIDIA/cccl/pull/6627
• Improve CMake package handling, add MSVC compat flags to libcudacxx's public interface. by @alliepiper in https://github.com/NVIDIA/cccl/pull/6791
• Fix arch related cuda::device:: APIs for nvhpc in CUDA mode by @davebayer in https://github.com/NVIDIA/cccl/pull/6829
• Implement the new tuning API for DeviceReduce by @bernhardmgruber in https://github.com/NVIDIA/cccl/pull/6544
• Replace internal assert with _CCCL_ASSERT in libcu++ by @davebayer in https://github.com/NVIDIA/cccl/pull/6825
• Implement std::gamma_distribution by @RAMitchell in https://github.com/NVIDIA/cccl/pull/6786
• Run scan benchmark for 2^32 elements by @bernhardmgruber in https://github.com/NVIDIA/cccl/pull/6834
• Remove upper bound on numba-cuda by @shwina in https://github.com/NVIDIA/cccl/pull/6835
• Use lit for cuda::arch_id and cuda::compute_capability tests by @davebayer in https://github.com/NVIDIA/cccl/pull/6775
• Extends DeviceScan tests in preparation for the warpspeed scan implementation by @elstehle in https://github.com/NVIDIA/cccl/pull/6836
• smoke test for all_of algorithm by @viralbhadeshiya in https://github.com/NVIDIA/cccl/pull/6828
• [CUB][device] Add a env-based overload of the device segmented reductions primitives by @rbourgeois33 in https://github.com/NVIDIA/cccl/pull/6674
• Avoid use of cccl namespace macros in cub by @davebayer in https://github.com/NVIDIA/cccl/pull/6844
• Beautify vector mismatch reporting by @bernhardmgruber in https://github.com/NVIDIA/cccl/pull/6837
• Implement std::lognormal_distribution by @RAMitchell in https://github.com/NVIDIA/cccl/pull/6789
• Implement std::weibull_distribution by @RAMitchell in https://github.com/NVIDIA/cccl/pull/6797
• [PTX] Add cp.async.bulk.dst.src.mbarrier::complete_tx::bytes.ignore_oob by @bernhardmgruber in https://github.com/NVIDIA/cccl/pull/6854
• Complex asinh accuracy refinement by @s-oboyle in https://github.com/NVIDIA/cccl/pull/6428
• Allow numpy struct types as initial value for Zipiterator inputs by @shwina in https://github.com/NVIDIA/cccl/pull/6861
• In test_device_segmented_scan_api change type from int to unsigned by @oleksandr-pavlyk in https://github.com/NVIDIA/cccl/pull/6868
• Add missing doc strings to support old CMake. by @alliepiper in https://github.com/NVIDIA/cccl/pull/6869
• avoid error adding pointer to reference in any_resource by @ericniebler in https://github.com/NVIDIA/cccl/pull/6875
• smoke test for adjacent_difference by @viralbhadeshiya in https://github.com/NVIDIA/cccl/pull/6872
• Bump minimum CMake to 3.18, add CI testing of public packages with it. by @alliepiper in https://github.com/NVIDIA/cccl/pull/6871
• [PTX] Regenerate by @bernhardmgruber in https://github.com/NVIDIA/cccl/pull/6859
• Implement std::poisson distribution by @RAMitchell in https://github.com/NVIDIA/cccl/pull/6748
• [libcu++] Add memory_pool header and correct legacy resources namespace by @pciolkosz in https://github.com/NVIDIA/cccl/pull/6852
• [cuda.compute]: Fix issue with get_dtype() not working anymore for pytorch arrays by @NaderAlAwar in https://github.com/NVIDIA/cccl/pull/6882
• [cuda.compute]: Add fast path to extract PyTorch array pointer by @NaderAlAwar in https://github.com/NVIDIA/cccl/pull/6884
• accommodate new behavior of clang's __builtin_structured_binding_size by @ericniebler in https://github.com/NVIDIA/cccl/pull/6888
• [libcu++] Don't require accessibility property on type erased wrappers by @pciolkosz in https://github.com/NVIDIA/cccl/pull/6851
• Move launch API from cudax to libcu++ by @pciolkosz in https://github.com/NVIDIA/cccl/pull/6667
• [libcu++] Fix minor version compatibility in 13.X by @pciolkosz in https://github.com/NVIDIA/cccl/pull/6895
• [libcu++] Leak static CUDA resources and add missing release on memory pool by @pciolkosz in https://github.com/NVIDIA/cccl/pull/6892
• Add limited RTX PRO 6000 coverage. by @alliepiper in https://github.com/NVIDIA/cccl/pull/6841
• Add quotes and error checking to devcontainer init. by @alliepiper in https://github.com/NVIDIA/cccl/pull/6886
• Update std::uniform_int_distribution by @RAMitchell in https://github.com/NVIDIA/cccl/pull/6799
• portability macro for checking whether an expression satisfies a concept in a _CCCL_REQUIRES_EXPR clause by @ericniebler in https://github.com/NVIDIA/cccl/pull/6890
• Add ptxas local memory usage warnings to cub builds by @davebayer in https://github.com/NVIDIA/cccl/pull/6838
• Pcg64 uint128 fallback implementation for MSVC by @RAMitchell in https://github.com/NVIDIA/cccl/pull/6746
• [cuda.compute] Add dependency on nvidia-nvvm by @shwina in https://github.com/NVIDIA/cccl/pull/6909
• Add missing OpKind docs entries by @ktaletsk in https://github.com/NVIDIA/cccl/pull/6910
• Fix overflow issue in histogram even benchmark when the number of bins exceeds what LevelT can represent by @NaderAlAwar in https://github.com/NVIDIA/cccl/pull/6908
• [libcu++] Add as_ref() to memory pool types by @pciolkosz in https://github.com/NVIDIA/cccl/pull/6900
• Fix exhaustive policy chain pruning test by @bernhardmgruber in https://github.com/NVIDIA/cccl/pull/6903
• Extend transform benchmarks to 2^32 elements by @bernhardmgruber in https://github.com/NVIDIA/cccl/pull/6920
• Implement std::cauchy_distribution by @RAMitchell in https://github.com/NVIDIA/cccl/pull/6787
• Implement std::extreme_value_distribution by @RAMitchell in https://github.com/NVIDIA/cccl/pull/6788
• Implement std::fisher_f_distribution by @RAMitchell in https://github.com/NVIDIA/cccl/pull/6857
• change concepts portability macros to avoid use of macro EXPAND by @ericniebler in https://github.com/NVIDIA/cccl/pull/6366
• Implement cuda::__all_arch_ids and cuda::__is_specific_arch by @davebayer in https://github.com/NVIDIA/cccl/pull/6916
• [libcu++] Rename device_transform back to launch_transform by @pciolkosz in https://github.com/NVIDIA/cccl/pull/6927
• Add unsupported compiler flag to .clangd by @ericniebler in https://github.com/NVIDIA/cccl/pull/6911
• Add an option to use CCCL from CTK for C2H by @bernhardmgruber in https://github.com/NVIDIA/cccl/pull/6848
• Avoid waring about missing braces for subobject by @miscco in https://github.com/NVIDIA/cccl/pull/6929
• Add missing nvrtc nv target archs by @davebayer in https://github.com/NVIDIA/cccl/pull/6880
• Make sure we actually use overflow builtins by @davebayer in https://github.com/NVIDIA/cccl/pull/6904
• Implement std::chi_squared_distribution by @RAMitchell in https://github.com/NVIDIA/cccl/pull/6856
• [libcu++] Static assert that resource is copyable in buffer constructors by @pciolkosz in https://github.com/NVIDIA/cccl/pull/6928
• Use vectorized transform kernel for sizeof(T) < 4 workloads of arity >1 on Hopper by @bernhardmgruber in https://github.com/NVIDIA/cccl/pull/6921
• port the trampoline_scheduler from stdexec to cudax::execution by @ericniebler in https://github.com/NVIDIA/cccl/pull/6894
• Properly specialize cub functions for __nv_bfloat16 by @miscco in https://github.com/NVIDIA/cccl/pull/6931
• [CUB] Fix mask types in block_radix_rank.cuh by @Aminsed in https://github.com/NVIDIA/cccl/pull/6189
• [CUB]: Use the new tuning API for nondeterministic reduce by @NaderAlAwar in https://github.com/NVIDIA/cccl/pull/6932
• clean up some allocator and memory utilities by @ericniebler in https://github.com/NVIDIA/cccl/pull/6939
• Unify operator handling in cuda.compute by @shwina in https://github.com/NVIDIA/cccl/pull/6938
• Use integer promotion for warp_reduce by @miscco in https://github.com/NVIDIA/cccl/pull/6819
• implement task_scheduler from C++26 ([exec.task.scheduler]) by @ericniebler in https://github.com/NVIDIA/cccl/pull/5975
• Implement std::negative_binomial_distribution by @RAMitchell in https://github.com/NVIDIA/cccl/pull/6879
• Implement std::student_t_distribution by @RAMitchell in https://github.com/NVIDIA/cccl/pull/6858
• Remove [[nodiscard]] from barrier's .arrive(...) method by @davebayer in https://github.com/NVIDIA/cccl/pull/6947
• Implement std::geometric_distribution by @RAMitchell in https://github.com/NVIDIA/cccl/pull/6924
• [cuda.compute] Refactor code for creating void* wrappers by @shwina in https://github.com/NVIDIA/cccl/pull/6941
• Expose not guaranteed determinism to reduce in cuda.compute by @NaderAlAwar in https://github.com/NVIDIA/cccl/pull/6926
• Make __cccl_is_floating_point_v consistent with __cccl_is_integer_v by @davebayer in https://github.com/NVIDIA/cccl/pull/6952
• Disable __builtin_structured_binding_size with nvcc by @davebayer in https://github.com/NVIDIA/cccl/pull/6961
• Provide thrust::find_if benchmark by @gonidelis in https://github.com/NVIDIA/cccl/pull/6956
• Remove all usage of old experimental MR macro by @pciolkosz in https://github.com/NVIDIA/cccl/pull/6962
• Fix cuda::std::abs for floating points by @davebayer in https://github.com/NVIDIA/cccl/pull/6958
• Expose <cuda/std/charconv> by @davebayer in https://github.com/NVIDIA/cccl/pull/6672
• Don't use __builtin_bswap128 during constant evaluation by @davebayer in https://github.com/NVIDIA/cccl/pull/6967
• Avoid using _CCCL_UNREACHABLE() unless it's necessary by @davebayer in https://github.com/NVIDIA/cccl/pull/6948
• Cleanups for random module by @RAMitchell in https://github.com/NVIDIA/cccl/pull/6951
• Remove need for hardcoded LevelT for histogram in c.parallel and cuda.compute by @NaderAlAwar in https://github.com/NVIDIA/cccl/pull/6915
• Use ublkcp/memcpy_async in transform when dtype size is not a power of two by @NaderAlAwar in https://github.com/NVIDIA/cccl/pull/6972
• Add internal cuda::__is_device_or_managed_memory by @fbusato in https://github.com/NVIDIA/cccl/pull/6918
• Improve and apply _CCCL_THROW by @fbusato in https://github.com/NVIDIA/cccl/pull/6684
• Define _CCCL_ASSERT_IMPL_HOST correctly for clang on Windows by @asmelko in https://github.com/NVIDIA/cccl/pull/6971
• Allow if consteval in device code with nvcc 13.1 by @davebayer in https://github.com/NVIDIA/cccl/pull/6902
• c.parallel: reuse CUB agent policies for histogram by @NaderAlAwar in https://github.com/NVIDIA/cccl/pull/6974
• [libcu++] Dynamically load CUDA library instead of using the runtime by @pciolkosz in https://github.com/NVIDIA/cccl/pull/6899
• [libcu++] Uncomment some tests and fix launch include after launch was moved to libcu++ by @pciolkosz in https://github.com/NVIDIA/cccl/pull/6966
• Re-enable using std::meow builtins by @davebayer in https://github.com/NVIDIA/cccl/pull/6978
• use cooperative_groups in execution::bulk to synchronize across thread blocks by @ericniebler in https://github.com/NVIDIA/cccl/pull/6992
• Provide Shared Memory mdspan and accessor by @fbusato in https://github.com/NVIDIA/cccl/pull/6703
• Add non-throwing overloads to is_pointer_accessible by @fbusato in https://github.com/NVIDIA/cccl/pull/6988
• [cuda.compute]: fix alignment not being set properly for gpu_struct types by @NaderAlAwar in https://github.com/NVIDIA/cccl/pull/6995
• Workaround for a potential bug in the driver related to TMA descriptor by @fbusato in https://github.com/NVIDIA/cccl/pull/6985
• do not introduce a pack in a structured binding with nvcc by @ericniebler in https://github.com/NVIDIA/cccl/pull/6994
• Extract environment boilerplate code from within the device interfaces to a separate header by @gonidelis in https://github.com/NVIDIA/cccl/pull/6622
• Remove _CCCL_HAS_CUDA_COMPILER() by @davebayer in https://github.com/NVIDIA/cccl/pull/6984
• [libcu++] Fix memory pool and buffer test issues on Windows by @pciolkosz in https://github.com/NVIDIA/cccl/pull/6993
• fix off-by-one error in the implementation of cuda::std::__tuple by @ericniebler in https://github.com/NVIDIA/cccl/pull/6996
• Small fixes around warpspeed scan by @bernhardmgruber in https://github.com/NVIDIA/cccl/pull/6998
• Remove <version> include by @davebayer in https://github.com/NVIDIA/cccl/pull/7001
• Upgrade GitHub Actions to latest versions by @salmanmkc in https://github.com/NVIDIA/cccl/pull/6991
• cuda.coop: Use cuda.core.experimental.Linker instead of internal numba-cuda _Linker by @shwina in https://github.com/NVIDIA/cccl/pull/7011
• Make c2h vector utils constexpr by @davebayer in https://github.com/NVIDIA/cccl/pull/7009
• Improves comments on decoupled look-back code example by @elstehle in https://github.com/NVIDIA/cccl/pull/7015
• Extract reduce_op_sync into a free function by @bernhardmgruber in https://github.com/NVIDIA/cccl/pull/7004
• Remove experimental namespace from cuda.core import by @NaderAlAwar in https://github.com/NVIDIA/cccl/pull/7022
• reexpress completion signature transform alias to make clangd happy by @ericniebler in https://github.com/NVIDIA/cccl/pull/7026
• Qualify call to __launch_impl in launch.h to avoid ambiguity errors by @ericniebler in https://github.com/NVIDIA/cccl/pull/7024
• Rework hierarchy levels by @davebayer in https://github.com/NVIDIA/cccl/pull/6957
• [CUB]: use vectorized kernel for triad and add benchmark for dtypes of size 2 by @NaderAlAwar in https://github.com/NVIDIA/cccl/pull/7019
• [libcu++] Fix synchronous resource adapter property passing by @pciolkosz in https://github.com/NVIDIA/cccl/pull/6976
• [libcu++] Remove _view from the shared memory getter name by @pciolkosz in https://github.com/NVIDIA/cccl/pull/6997
• [thrust] Ignore CUDA free errors in thrust memory resource by @pciolkosz in https://github.com/NVIDIA/cccl/pull/7002
• [libcu++] Correctly handle extended lambda in cuda::launch by @pciolkosz in https://github.com/NVIDIA/cccl/pull/6987
• Use <stdexcept> header unconditionally by @fbusato in https://github.com/NVIDIA/cccl/pull/7028
• Error out when nvrtcc cannot parse cuda_thread_count by @bernhardmgruber in https://github.com/NVIDIA/cccl/pull/7035
• Allow all public headers to be included with host compilers only by @davebayer in https://github.com/NVIDIA/cccl/pull/7012
• [cuda.compute]: Fixes and updates to benchmarks by @shwina in https://github.com/NVIDIA/cccl/pull/6999
• Support operations with side-effects (state) in cuda.compute by @shwina in https://github.com/NVIDIA/cccl/pull/7008
• Fix cuda::memcpy async edge cases and add more tests by @bernhardmgruber in https://github.com/NVIDIA/cccl/pull/6608
• Explicitly set CCCL_TOPLEVEL_PROJECT to OFF when needed by @KyleFromNVIDIA in https://github.com/NVIDIA/cccl/pull/7016
• [libcu++] Add explicit alignment specification in buffer by @pciolkosz in https://github.com/NVIDIA/cccl/pull/7005
• Use the sccache-dist build cluster for RAPIDS CI jobs by @trxcllnt in https://github.com/NVIDIA/cccl/pull/7014
• tidy up the primitive variant type used by cudax::execution by @ericniebler in https://github.com/NVIDIA/cccl/pull/7029
• Fix docs by @gevtushenko in https://github.com/NVIDIA/cccl/pull/7052
• Disable LDL/STL checks, for failures seen with NVRTC 13.1 by @shwina in https://github.com/NVIDIA/cccl/pull/7054
• Enhance DLPack compatibility by @fbusato in https://github.com/NVIDIA/cccl/pull/7045
• Support lambdas as operators in cuda.compute by @shwina in https://github.com/NVIDIA/cccl/pull/7058
• [libcu++] Make kernel_config member private and allow it in hierarchy queries by @pciolkosz in https://github.com/NVIDIA/cccl/pull/7034
• [libcu++] Remove mentions of cuda/event header from docs by @pciolkosz in https://github.com/NVIDIA/cccl/pull/7066
• [BUG] use references for mdspan internal methods by @fbusato in https://github.com/NVIDIA/cccl/pull/7059
• Avoid invalid compiler warning with VS2026 by @miscco in https://github.com/NVIDIA/cccl/pull/7077
• Avoid compiler issue with MSVC _CCCL_UNREACHABLE by @miscco in https://github.com/NVIDIA/cccl/pull/7080
• cuda.compute: Allow multiple uses of the same function in single compilation by @shwina in https://github.com/NVIDIA/cccl/pull/7072
• Refactor c2h generator to ensure teardown before main exits by @bernhardmgruber in https://github.com/NVIDIA/cccl/pull/7067
• Remove cumlprims_mg from RAPIDS workflows/devcontainers by @bdice in https://github.com/NVIDIA/cccl/pull/7082
• [DOCS] Clarifies DeviceTopK docs that inputs and output ranges may not overlap by @elstehle in https://github.com/NVIDIA/cccl/pull/7078
• Enhance RDC detection and add _CCCL_HAS_DEVICE_RUNTIME() macro by @davebayer in https://github.com/NVIDIA/cccl/pull/7049
• Expand warning suppression for braces around subobject by @miscco in https://github.com/NVIDIA/cccl/pull/7087
• [STF] Document how to enable assertions by @caugonnet in https://github.com/NVIDIA/cccl/pull/7084
• Simplify cuda::host_launch API by @davebayer in https://github.com/NVIDIA/cccl/pull/6689
• Improvements to cuda.compute documentation by @shwina in https://github.com/NVIDIA/cccl/pull/7061
• [libcu++] Add tests for some buffer members and alignment passing by @pciolkosz in https://github.com/NVIDIA/cccl/pull/7055
• [libcu++] Fix driver api test after curand changes by @pciolkosz in https://github.com/NVIDIA/cccl/pull/7095
• Add DeviceTransform to device wide CUB docs by @bernhardmgruber in https://github.com/NVIDIA/cccl/pull/7101
• Fix incorrect if else logic in fmax by @miscco in https://github.com/NVIDIA/cccl/pull/7107
• Add -device-type128 flags only once by @davebayer in https://github.com/NVIDIA/cccl/pull/7100
• [libcu++] Check if managed pools are accessible in is_pointer_accessible test by @pciolkosz in https://github.com/NVIDIA/cccl/pull/7096
• Fix calculation of necessary bits in feistel projection by @miscco in https://github.com/NVIDIA/cccl/pull/7098
• Fix deferred annotations handling in gpu_struct by @shwina in https://github.com/NVIDIA/cccl/pull/7121
• Use cudaMemcpyDefault for trivial copies by @bdice in https://github.com/NVIDIA/cccl/pull/7006
• Disable NVHPC builds for pull request CI by @miscco in https://github.com/NVIDIA/cccl/pull/7135
• Generator for prologue/epilogue by @davebayer in https://github.com/NVIDIA/cccl/pull/7099
• Refactor mdspan cuda::std::__detectably_invalid by @fbusato in https://github.com/NVIDIA/cccl/pull/6733
• Fix nvrtcc minimum arch for __float128 support by @davebayer in https://github.com/NVIDIA/cccl/pull/7119
• Disable cudax with msvc in CI for now by @pciolkosz in https://github.com/NVIDIA/cccl/pull/7139
• Simplify namespace definitions by @davebayer in https://github.com/NVIDIA/cccl/pull/7104
• Move DLPack include to separate file by @davebayer in https://github.com/NVIDIA/cccl/pull/7108
• Replace and deprecate compute_capability::major() and compute_capability::minor() by @davebayer in https://github.com/NVIDIA/cccl/pull/7118
• Disable reference_wrapper test for VS2026 by @miscco in https://github.com/NVIDIA/cccl/pull/7088
• Clean up hierarchy by @davebayer in https://github.com/NVIDIA/cccl/pull/7023
• Implement new tuning API arch dispatching by @bernhardmgruber in https://github.com/NVIDIA/cccl/pull/7093
• Improve std:: builtin handling with nvrtc by @davebayer in https://github.com/NVIDIA/cccl/pull/7131
• libcu++: silence msvc+nvcc12.9 warning plaguing c.parallel. by @griwes in https://github.com/NVIDIA/cccl/pull/7144
• Implement cub::DeviceFind::FindIf by @gonidelis in https://github.com/NVIDIA/cccl/pull/2405
• Fix/modernize thrust examples by @Flawxd in https://github.com/NVIDIA/cccl/pull/7094
• Modularize chrono by @miscco in https://github.com/NVIDIA/cccl/pull/6671
• Those are unused internal traits by @miscco in https://github.com/NVIDIA/cccl/pull/7148
• Implement ranges::reverse_view by @miscco in https://github.com/NVIDIA/cccl/pull/6751
• Fixes for shuffle_iterator by @RAMitchell in https://github.com/NVIDIA/cccl/pull/7130
• [STF] Use execution places without STF contexts by @caugonnet in https://github.com/NVIDIA/cccl/pull/7149
• Revert nested namespace change to <nv/target> by @wmaxey in https://github.com/NVIDIA/cccl/pull/7151
• Replace internal uses of thrust::tuple with cuda::std::tuple by @miscco in https://github.com/NVIDIA/cccl/pull/6629
• Add Android-specific assert handling in __cccl/assert.h by @fbusato in https://github.com/NVIDIA/cccl/pull/7156
• Fix make_tma_descriptor() unit test by @fbusato in https://github.com/NVIDIA/cccl/pull/7152
• Rename new tuning API policies and fix MSVC warning by @bernhardmgruber in https://github.com/NVIDIA/cccl/pull/7103
• Align local vector storage arrays in vec transform by @bernhardmgruber in https://github.com/NVIDIA/cccl/pull/7162
• Try and avoid GCC-15 warning about expected ) by @miscco in https://github.com/NVIDIA/cccl/pull/7166
• Fix build issues with documentation by @miscco in https://github.com/NVIDIA/cccl/pull/7122
• [DOCS] Improves docs for DeviceTopK, clarifying that inputs and output ranges must not overlap by @elstehle in https://github.com/NVIDIA/cccl/pull/7086
• Test passing a custom policy to DispatchRadixSort by @bernhardmgruber in https://github.com/NVIDIA/cccl/pull/7170
• Avoid benign overflow in __calloc_device by @miscco in https://github.com/NVIDIA/cccl/pull/7176
• Fixes for thrust::shuffle by @RAMitchell in https://github.com/NVIDIA/cccl/pull/7172
• cub, c.parallel: {lower,upper}_bound by @griwes in https://github.com/NVIDIA/cccl/pull/7007
• Initial nvrtcc implementation by @davebayer in https://github.com/NVIDIA/cccl/pull/7051
• Implement comparison operators for thrust::reference and thrust::pointer by @miscco in https://github.com/NVIDIA/cccl/pull/7190
• Agent Updates by @alliepiper in https://github.com/NVIDIA/cccl/pull/7194
• Add missing CUB_RUNTIME_FUNCTION annotations. by @alliepiper in https://github.com/NVIDIA/cccl/pull/7195
• Define methods for test ranges by @miscco in https://github.com/NVIDIA/cccl/pull/7220
• Fix noexcept specification of extreme_value_distribution by @miscco in https://github.com/NVIDIA/cccl/pull/7219
• Drop thrust::detail::is_commutative by @miscco in https://github.com/NVIDIA/cccl/pull/7218
• Try and work around NVHPC issue with is_metafunction_defined by @miscco in https://github.com/NVIDIA/cccl/pull/7217
• Reenable MSVC cudax CI by @miscco in https://github.com/NVIDIA/cccl/pull/7221
• Add support for [[lifetimebound]] by @fbusato in https://github.com/NVIDIA/cccl/pull/7155
• [cccl.c] Use function try blocks by @davebayer in https://github.com/NVIDIA/cccl/pull/7236
• Do not try to run catch2 tests with nvrtc by @miscco in https://github.com/NVIDIA/cccl/pull/7242
• Add .branch_notes. by @alliepiper in https://github.com/NVIDIA/cccl/pull/7238
• Make cuda.compute importable in a CPU-only environment by @shwina in https://github.com/NVIDIA/cccl/pull/7171
• Drop libcudacxx ABI Evolution clause by @bernhardmgruber in https://github.com/NVIDIA/cccl/pull/7247
• [CI] MSVC sccache auth -> file by @alliepiper in https://github.com/NVIDIA/cccl/pull/7257
• Implement the new tuning API for DeviceTransform by @bernhardmgruber in https://github.com/NVIDIA/cccl/pull/6914
• Refactor some bits of DeviceRadixSort by @bernhardmgruber in https://github.com/NVIDIA/cccl/pull/7193
• Drop leftover code after tuning API migration by @bernhardmgruber in https://github.com/NVIDIA/cccl/pull/7264
• Fix extracting CUDA stream in cub::DeviceTransform by @bernhardmgruber in https://github.com/NVIDIA/cccl/pull/7239
• Run build/test commands with 5h30m timeout on CI. by @alliepiper in https://github.com/NVIDIA/cccl/pull/7213
• Change the order of conditions in cuda::barrier by @davebayer in https://github.com/NVIDIA/cccl/pull/7259
• Don't run CPU-only import test if the wheel artifact doesn't exist by @shwina in https://github.com/NVIDIA/cccl/pull/7270
• Test passing more stream types to cub::DeviceTransform by @bernhardmgruber in https://github.com/NVIDIA/cccl/pull/7278
• Add versionadded directives to all public API functions by @cliffburdick in https://github.com/NVIDIA/cccl/pull/7215
• Refactor cub::DeviceRadixSort by @bernhardmgruber in https://github.com/NVIDIA/cccl/pull/7282
• [FEA]: Add DevEx/Infra ticket templates by @alliepiper in https://github.com/NVIDIA/cccl/pull/7261
• Fix __query_or CPO by @miscco in https://github.com/NVIDIA/cccl/pull/7266
• Test passing a custom policy to DispatchUniqueByKey by @bernhardmgruber in https://github.com/NVIDIA/cccl/pull/7296
• Test passing a custom policy to DispatchSelectIf by @bernhardmgruber in https://github.com/NVIDIA/cccl/pull/7294
• Test passing a custom policy to DispatchHistogram by @bernhardmgruber in https://github.com/NVIDIA/cccl/pull/7288
• Fix is_address_from for cluster_shared for pre-sm_90 by @davebayer in https://github.com/NVIDIA/cccl/pull/7245
• Test passing a custom policy to DispatchThreeWayPartitionIf by @bernhardmgruber in https://github.com/NVIDIA/cccl/pull/7295
• cuda.compute: Consolidate caching logic across all algorithms by @shwina in https://github.com/NVIDIA/cccl/pull/7281
• Test passing a custom policy to DispatchAdjacentDifference, DispatchMergeSort, DispatchScan, DispatchBatchMemcpy by @bernhardmgruber in https://github.com/NVIDIA/cccl/pull/7289
• Test passing a custom policy to DispatchSegmentedSort by @bernhardmgruber in https://github.com/NVIDIA/cccl/pull/7307
• Test passing a custom policy to DispatchSegmentedRadixSort by @bernhardmgruber in https://github.com/NVIDIA/cccl/pull/7308
• Test passing a custom policy to DispatchSegmentedReduce by @bernhardmgruber in https://github.com/NVIDIA/cccl/pull/7311
• Add accessor methods to shared_resource<T> by @bdice in https://github.com/NVIDIA/cccl/pull/7315
• cub, c.parallel: change {lower,upper}_bound to return indices. by @griwes in https://github.com/NVIDIA/cccl/pull/7320
• [nvrtcc] Add __NVRTCC_USE_NVRTC__ macro to nvrtcc by @davebayer in https://github.com/NVIDIA/cccl/pull/7293
• Remove fmtlib from CCCL by @davebayer in https://github.com/NVIDIA/cccl/pull/7300
• Fix clang warning about missing braces again by @miscco in https://github.com/NVIDIA/cccl/pull/7302
• Test passing a custom policy to DispatchReduceByKey by @bernhardmgruber in https://github.com/NVIDIA/cccl/pull/7310
• Fix missing newline for spdx/pragma once by @msarahan in https://github.com/NVIDIA/cccl/pull/7306
• mdspan to DLPack by @fbusato in https://github.com/NVIDIA/cccl/pull/7027
• Test passing a custom policy to DeviceRleDispatch by @bernhardmgruber in https://github.com/NVIDIA/cccl/pull/7314
• Skip checking build prereqs if installing by @wmaxey in https://github.com/NVIDIA/cccl/pull/7316
• Implement the new tuning API for DeviceRadixSort by @bernhardmgruber in https://github.com/NVIDIA/cccl/pull/6767
• Add cuda.compute APIs for upper_bound and lower_bound by @shwina in https://github.com/NVIDIA/cccl/pull/7250
• cuda.compute: Fix deferred annotations handling in signature_from_annotations by @shwina in https://github.com/NVIDIA/cccl/pull/7321
• Test passing a custom policy to DispatchScanByKey by @bernhardmgruber in https://github.com/NVIDIA/cccl/pull/7309
• Fixup missed feedback on [#7311] by @bernhardmgruber in https://github.com/NVIDIA/cccl/pull/7323
• Enhance internal vector type utilities by @fbusato in https://github.com/NVIDIA/cccl/pull/7327
• Fix MSVC version detection by @miscco in https://github.com/NVIDIA/cccl/pull/7305
• Implement parallel cuda::std::reduce by @miscco in https://github.com/NVIDIA/cccl/pull/6777
• Remove _CCCL_HAS_INCLUDE by @davebayer in https://github.com/NVIDIA/cccl/pull/7304
• Fix narrow conversion in __is_valid_address_range when compling on 32-bit systems by @davebayer in https://github.com/NVIDIA/cccl/pull/7333
• [nvrtcc] Add nvrtcc dependency by @davebayer in https://github.com/NVIDIA/cccl/pull/7287
• Fix double destroy in vector by @miscco in https://github.com/NVIDIA/cccl/pull/7331
• Optimize cuda::add_overflow for unsigned types by @davebayer in https://github.com/NVIDIA/cccl/pull/7340
• [STF] Make data place more extensible by @caugonnet in https://github.com/NVIDIA/cccl/pull/7252
• Formatters for cuda::arch_id and cuda::compute_capability by @davebayer in https://github.com/NVIDIA/cccl/pull/7335
• Drop most parts of thrust::allocator_traits by @miscco in https://github.com/NVIDIA/cccl/pull/7286
• [CI] Re-enable RAPIDS builds in PRs by @alliepiper in https://github.com/NVIDIA/cccl/pull/7255
• Use arch dispatch workaround on GCC 8-9 as well by @bernhardmgruber in https://github.com/NVIDIA/cccl/pull/7349
• Drop partial specialization in friend functions of fast_mod_div by @miscco in https://github.com/NVIDIA/cccl/pull/7348
• Add blackduck-sca.yml by @alliepiper in https://github.com/NVIDIA/cccl/pull/7360
• [cccl.c]: Disable SASS check for merge_sort pairs by @NaderAlAwar in https://github.com/NVIDIA/cccl/pull/7357
• Assert before terminating on throw in device code by @davebayer in https://github.com/NVIDIA/cccl/pull/7358
• DLPack to mdspan by @fbusato in https://github.com/NVIDIA/cccl/pull/7047
• Test VerifyCodegen only with latest CTK by @davebayer in https://github.com/NVIDIA/cccl/pull/7351
• Ensure device_find_if works with non-default-constructible types by @miscco in https://github.com/NVIDIA/cccl/pull/7337
• Fix _CCCL_THROW in dlpack_to_mdspan by @fbusato in https://github.com/NVIDIA/cccl/pull/7363
• Expose CUDA vector type traits by @fbusato in https://github.com/NVIDIA/cccl/pull/7364
• Add some ramblings about symbol visibility by @miscco in https://github.com/NVIDIA/cccl/pull/6114
• Fix handling of boolean types in cuda.compute by @shwina in https://github.com/NVIDIA/cccl/pull/7389
• Decouple numba-cuda from type system and other internals by @shwina in https://github.com/NVIDIA/cccl/pull/7342
• Fix DeviceTransform docs by @bernhardmgruber in https://github.com/NVIDIA/cccl/pull/7392
• [libcu++] Add runtime check if memory pools are supported by @pciolkosz in https://github.com/NVIDIA/cccl/pull/7339
• Pass -std flag from CLI to cmake in c.parallel build scripts by @NaderAlAwar in https://github.com/NVIDIA/cccl/pull/7394
• Migrate cuco HLL by @srinivasyadav18 in https://github.com/NVIDIA/cccl/pull/6666
• Add CTK 13.1 CI jobs, devcontainers. by @alliepiper in https://github.com/NVIDIA/cccl/pull/6887
• make the abi of __basic_any compatible between c++17 and c++20 by @ericniebler in https://github.com/NVIDIA/cccl/pull/7401
• Enforce supported long double format by @davebayer in https://github.com/NVIDIA/cccl/pull/7345
• Implement passing stream and memory resource to execution policies by @miscco in https://github.com/NVIDIA/cccl/pull/7299
• Reenable overflow builtins with nvc++ 26.1+ by @davebayer in https://github.com/NVIDIA/cccl/pull/7414
• Implement the new tuning API for DeviceSegmentedReduce by @bernhardmgruber in https://github.com/NVIDIA/cccl/pull/7334
• part deux: make the abi of __basic_any compatible between c++17 and c++20 by @ericniebler in https://github.com/NVIDIA/cccl/pull/7405
• Move host stdlib wrappers to <cuda/std/__host_stdlib> directory by @davebayer in https://github.com/NVIDIA/cccl/pull/7411
• Improves error handling in CUB algorithms using thrust::triple_chevron calls by @elstehle in https://github.com/NVIDIA/cccl/pull/7415
• [DOCS] Clarifies that thrust::reduce_by_key actually selects the last item of a range of consecutively equal keys by @elstehle in https://github.com/NVIDIA/cccl/pull/7408
• fix determinism rejection logic for scan by @srinivasyadav18 in https://github.com/NVIDIA/cccl/pull/7133
• Upgrade GitHub Actions for Node 24 compatibility by @salmanmkc in https://github.com/NVIDIA/cccl/pull/6990
• [libcu++] Fix handle-type mask checks in memory pool tests. by @pciolkosz in https://github.com/NVIDIA/cccl/pull/7428
• Avoid deallocate throwing by @gonidelis in https://github.com/NVIDIA/cccl/pull/7233
• Drop accidental [[nodiscard]] on constructor by @miscco in https://github.com/NVIDIA/cccl/pull/7413
• Use SPDX license headers in thrust/thrust/system/detail/generic by @bernhardmgruber in https://github.com/NVIDIA/cccl/pull/7422
• Use SPDX license headers in thrust/thrust/detail by @bernhardmgruber in https://github.com/NVIDIA/cccl/pull/7423
• Update Catch2 to 3.12 by @bernhardmgruber in https://github.com/NVIDIA/cccl/pull/6067
• Add DeviceTransform benchmarks from pytorch by @bernhardmgruber in https://github.com/NVIDIA/cccl/pull/7391
• Refactor cub::ThreadLoad by @bernhardmgruber in https://github.com/NVIDIA/cccl/pull/7419
• cuda.compute: Don't attempt to set host_advance by @shwina in https://github.com/NVIDIA/cccl/pull/7425
• Use SPDX license headers in thrust/system/cpp by @bernhardmgruber in https://github.com/NVIDIA/cccl/pull/7430
• Use SPDX license headers in thrust/random by @bernhardmgruber in https://github.com/NVIDIA/cccl/pull/7435
• Use SPDX license headers in thrust/system/cuda by @bernhardmgruber in https://github.com/NVIDIA/cccl/pull/7432
• Use SPDX license headers in thrust/iterator by @bernhardmgruber in https://github.com/NVIDIA/cccl/pull/7433
• Use SPDX license headers some thrust files 2/2 by @bernhardmgruber in https://github.com/NVIDIA/cccl/pull/7437
• Use SPDX license headers in thrust/system/detail/sequential by @bernhardmgruber in https://github.com/NVIDIA/cccl/pull/7431
• Use SPDX license headers some thrust files 1/2 by @bernhardmgruber in https://github.com/NVIDIA/cccl/pull/7436
• Optimize cuda::sub_overflow by @davebayer in https://github.com/NVIDIA/cccl/pull/7344
• [c.parallel]: migrate transform to use jit templates instead of string based implementations by @NaderAlAwar in https://github.com/NVIDIA/cccl/pull/7399
• Fix build against libc++ by @miscco in https://github.com/NVIDIA/cccl/pull/7448
• cuda.compute: Fix struct comparison (ordering matters) by @shwina in https://github.com/NVIDIA/cccl/pull/7451
• Disable batch benchmarks for DeviceTransform by @bernhardmgruber in https://github.com/NVIDIA/cccl/pull/7450
• Initial version of DeviceSegmentedTopk for fixed-size segments by @elstehle in https://github.com/NVIDIA/cccl/pull/6980
• Document random module by @RAMitchell in https://github.com/NVIDIA/cccl/pull/7412
• Check for __cpp_xxx value, not definition by @davebayer in https://github.com/NVIDIA/cccl/pull/4811
• Fix missing c2h symbol when compiling with clang-cuda by @davebayer in https://github.com/NVIDIA/cccl/pull/7454
• Optimize cuda::add_overflow for signed types by @davebayer in https://github.com/NVIDIA/cccl/pull/7343
• Implement parallel cuda::std::transform by @miscco in https://github.com/NVIDIA/cccl/pull/7395
• [cuda.coop]: add device-side coop.warp.sum benchmark with pynvbench by @NaderAlAwar in https://github.com/NVIDIA/cccl/pull/6846
• Replace typedef with using by @davebayer in https://github.com/NVIDIA/cccl/pull/7271
• Refactor cub::ThreadStore by @bernhardmgruber in https://github.com/NVIDIA/cccl/pull/7418
• Fix non default constructible input types test for cub::FindIf by @gonidelis in https://github.com/NVIDIA/cccl/pull/7447
• Add versionadded annotations to CUB public APIs by @cliffburdick in https://github.com/NVIDIA/cccl/pull/7406
• Fix random doc formatting by @RAMitchell in https://github.com/NVIDIA/cccl/pull/7492
• Include ninja, ctest, and sccache logs in build artifacts by @trxcllnt in https://github.com/NVIDIA/cccl/pull/7487
• Suppress MSVC-specific warnings on linux. by @alliepiper in https://github.com/NVIDIA/cccl/pull/7462
• Drop usage of pickle by @miscco in https://github.com/NVIDIA/cccl/pull/7491
• Add cub::DeviceTransform N->M API entrypoint by @bernhardmgruber in https://github.com/NVIDIA/cccl/pull/7473
• Improve reduce implementation by @miscco in https://github.com/NVIDIA/cccl/pull/7493
• Minor libcu++ lit config improvements by @trxcllnt in https://github.com/NVIDIA/cccl/pull/7486
• Implement parallel cuda::std::replace by @miscco in https://github.com/NVIDIA/cccl/pull/7407
• cuda.compute: improve caching performance by not relying on isinstance() checks for protocols by @shwina in https://github.com/NVIDIA/cccl/pull/7501
• Retry failed image pulls 10 times by @trxcllnt in https://github.com/NVIDIA/cccl/pull/7488
• Fix DeviceReduce env test for rfa by @bernhardmgruber in https://github.com/NVIDIA/cccl/pull/7481
• Pass PARALLEL_LEVEL to cmake --build in ci/build_stdpar.sh by @trxcllnt in https://github.com/NVIDIA/cccl/pull/7483
• cuda.compute: Use native CCCL.c support for stateful ops by @shwina in https://github.com/NVIDIA/cccl/pull/7500
• [STF] Add exec_place_guard RAII helper for scoped exec_place activation by @caugonnet in https://github.com/NVIDIA/cccl/pull/7434
• Implement parallel cuda::std::replace_copy by @miscco in https://github.com/NVIDIA/cccl/pull/7410
• Provide Operator Properties by @fbusato in https://github.com/NVIDIA/cccl/pull/7240
• Implement parallel cuda::std::generate by @miscco in https://github.com/NVIDIA/cccl/pull/7416
• [STF] Fix equality operators for places by @caugonnet in https://github.com/NVIDIA/cccl/pull/7494
• Implement parallel cuda::std::count by @miscco in https://github.com/NVIDIA/cccl/pull/7382
• Clarify Docker install step for WSL in .devcontainer README by @acosmicflamingo in https://github.com/NVIDIA/cccl/pull/7516
• Restrict to numba-cuda less than 0.27 by @shwina in https://github.com/NVIDIA/cccl/pull/7529
• Fix caching of functions referencing numpy ufuncs (or any "dotted" functions) by @shwina in https://github.com/NVIDIA/cccl/pull/7535
• Add docs for cccl-runtime 3.2 additions and cccl-runtime landing page by @pciolkosz in https://github.com/NVIDIA/cccl/pull/7489
• Tweak random docs by @RAMitchell in https://github.com/NVIDIA/cccl/pull/7534
• Use modern __syncthreads_or primitive by @gonidelis in https://github.com/NVIDIA/cccl/pull/7509
• Fix identity_element tests by @miscco in https://github.com/NVIDIA/cccl/pull/7526
• [STF] Use relaxed capture mode by @caugonnet in https://github.com/NVIDIA/cccl/pull/7566
• Remove recursion from __internal_is_address_from by @dkolsen-pgi in https://github.com/NVIDIA/cccl/pull/7561
• Fix operator identity and absorbing element for char by @bernhardmgruber in https://github.com/NVIDIA/cccl/pull/7568
• Rewrite iterators to not depend on Numba by @shwina in https://github.com/NVIDIA/cccl/pull/7441
• [Backport branch/3.3.x] Fix ranges_overlap for nvc++ -cuda by @github-actions[bot] in https://github.com/NVIDIA/cccl/pull/7599
• [Backport branch/3.3.x] Fix cuda::device::current_arch_id by @github-actions[bot] in https://github.com/NVIDIA/cccl/pull/7602
• [Backport branch/3.3.x] Fix cuda::barrier missing accounting of results in try_wait by @github-actions[bot] in https://github.com/NVIDIA/cccl/pull/7635
• [Backport branch/3.3.x] Check for _GLIBCXX_USE_CXX11_ABI only when compiling with libstdc++ by @github-actions[bot] in https://github.com/NVIDIA/cccl/pull/7631

New Contributors

• @miyanyan made their first contribution in https://github.com/NVIDIA/cccl/pull/6773
• @Pansysk75 made their first contribution in https://github.com/NVIDIA/cccl/pull/6572
• @rbourgeois33 made their first contribution in https://github.com/NVIDIA/cccl/pull/6674
• @ktaletsk made their first contribution in https://github.com/NVIDIA/cccl/pull/6910
• @asmelko made their first contribution in https://github.com/NVIDIA/cccl/pull/6971
• @salmanmkc made their first contribution in https://github.com/NVIDIA/cccl/pull/6991
• @KyleFromNVIDIA made their first contribution in https://github.com/NVIDIA/cccl/pull/7016
• @Flawxd made their first contribution in https://github.com/NVIDIA/cccl/pull/7094
• @msarahan made their first contribution in https://github.com/NVIDIA/cccl/pull/7306
• @acosmicflamingo made their first contribution in https://github.com/NVIDIA/cccl/pull/7516

Full Changelog: https://github.com/NVIDIA/cccl/compare/v3.3.0.dev...v3.3.0