The compute-oriented DGX-2 server launched in early 2018 is getting an upgrade consisting of more powerful Xeon CPUs and higher-clocked Tesla V100 GPUs. The Tesla V100 cards come with a total of 512 GB of HBM2 memory and are interconnected through special NVLinks, allowing them to act as a single GPU with a massive throughput of 2.1 PetaFLOPS.
https://www.notebookcheck.net/Nvidia-announces-DGX-2H-compute-server-with-upgraded-Tesla-V100-GPUs.366026.0.html