All samples make use of the driver API.
This sample enumerates the properties of the CUDA devices present in the system.
This sample measures host to device and device to host copy bandwidth for pageable, page-locked and write-combined memory of transfer sizes 3KB, 15KB, 15MB and 100MB, and outputs them in CSV format.
This sample jit-in-time compiles a .ptx and outputs error log and info log.
This sample uses zero copy to map a host pointer to a device pointer so that kernels can read and write directly to pinned system memory.
This sample uses async API, dynamic ptx version selection, and constant and shared memory to add two vectors of float.
This sample uses multiple streams to exploit the HyperQ technology.
This sample uses multiple devices to parallelize computation.