libcu++, the NVIDIA C++ Standard Library, provides a C++ Standard Library for your entire system which can be used in and between CPU and GPU codes. The NVIDIA C++ Standard Library is an open source project.
Version 1.4.0 of libcu++ is a major release providing several feature enhancements and bug fixes. It adds support for the following: <cuda/std/complex>, NVCC + MSVC support for <cuda/std/tuple>, and backports of C++20 <cuda/std/chrono> and C++17 <cuda/std/type_traits> features to C++14.
Other enhancements include improved and reorganized documentation, atomics decoupled from host Standard Library in MSVC, and revamped examples and benchmarks.
Additional information, examples and documentation can be found below.