The CUDA 6 platform features unified memory and a number of other new features that make parallel programming easier than ever and enable software developers to dramatically decrease the time and effort required to accelerate their scientific, engineering, enterprise and other applications.