NVLink is a communications protocol developed by Nvidia. NVLink specifies a point-to-point connection between a CPU and a GPU and also between a GPU and another GPU rated at 80 GB/s. NVLink products introduced to date focus on the high-performance application space.
On 5 April 2016, Nvidia announced that NVLink will be implemented in the forthcoming Pascal-microarchitecture-based GP100 GPU, which will be used in, for example, Nvidia Tesla P100 products.
The US Department of Energy contracted Nvidia and IBM to build two supercomputers named "Summit" and "Sierra", which will use NVLink for the node interconnects, while a variant of InfiniBand will be used for the system interconnects. These systems will combine Nvidia's Volta architecture with the POWER9 family of CPUs.