.. _libcudacxx-extended-api-asynchronous-operations: Asynchronous Operations ----------------------- .. toctree:: :hidden: :maxdepth: 1 asynchronous_operations/memcpy_async_tx asynchronous_operations/memcpy_async .. list-table:: :widths: 25 45 30 30 :header-rows: 1 * - **Header** - **Content** - **CCCL Availability** - **CUDA Toolkit Availability** * - :ref:`cuda::memcpy_async ` - Asynchronously copies one range to another - libcu++ 1.1.0 / CCCL 2.0.0 - CUDA 11.0 * - :ref:`cuda::memcpy_async_tx ` - Asynchronously copies one range to another with manual transaction accounting - libcu++ 1.2.0 / CCCL 2.0.0 - CUDA 11.1 .. note:: **Asynchronous operations** like `memcpy_async ` are non-blocking operations performed as-if by a new thread of execution.