CPU (central processing unit). The central unit in a computer responsible for calculations and for controlling or supervising other parts of the computer. The CPU performs logical and floating point operations on data held in the computer memory.

GPU (graphics processing unit). Programmable chip originally intended for graphics rendering. The highly parallel structure of a GPU makes them more effective than general-purpose CPUs for algorithms where processing of large blocks of data is done in parallel.

Core. A single independent computational unit within a CPU or GPU chip. CPU and GPU cores are not equivalent to each other; GPU cores perform specialized operations whereas CPU cores are designed for general-purpose programs.

CUDA®. A parallel computing technology from NVIDIA® that consists of a parallel computing architecture and developer tools, libraries, and programming directives for GPU computing.

Device. A hardware card containing the GPU and its associated memory.

Host. The CPU and system memory.

Kernel. Code written for execution on the GPU. Kernels are functions that can run on a large number of threads. The parallelism arises from each thread independently running the same program on different data.