The programming support for NVIDIA GPUs in Julia is built on the CUDA toolkit, and aims to be as full-featured and offer the same performance as working with GPUs in CUDA C. It is a mature toolchain that has been under development since 2014 and is supported out-of-the-box on all current versions of Julia. The toolchain is composed of several Julia packages, ecah with separate responsibilities: CUDAdrv.jl: interfacing with the CUDA driver library; CUDAnative.jl: GPU kernel programming support; CuArrays.jl: array programming interface