Microsoft and its public cloud, Azure, is expanded with the help of Nvidia to provide users with capability of N-series GPUs remotely. For this, virtual machines are in place over a fast RDMA network.
Users can expect 35% faster instances compared to Azure D series. It is aimed at applications needing faster CPU, higher memory and demanding high local disk performance. The good news is that DV2 announcement has also resulted in reduction of prices of D-Series.
The second instance to be announced by MS is GPU-backed N-series. This one will provide users access to NVIDIA Tesla K80 GPU accelerator used mainly for HPC. This instance will also provide enterprise workflows with NVIDIA GRID 2.0 technology for virtualized graphics.
On the other hand, NVIDIA CEO Jen-Hsun Huang released a video sharing more details about cloud-based GPUs. In the video he said:
It takes an enormous amount of innovation in order to put GPUs into the cloud. These are processors that are intensely high-performance but want to be very close to the user. Through very important innovations, we’ve now made it possible for these powerful GPUs to be included in the cloud. The first is the intense energy-efficiency focus that is required because data centers are really large and the cost to manage the datacenter is very high.
Rest of the details can be found in the video above.