Microsoft today announced the general availability of the Azure N-Series of Azure Virtual Machines as of December 1, 2016. Azure N-Series virtual will be first available in East US, West Europe, and South East Asia at GA joining the current South Central US region. N-Series instances are enabled with NVIDIA’s GPUs to allow you to run GPU-accelerated workloads and visualize them. N-Series VMs are split into two categories: NC-Series and NV-Series.
Azure NC virtual machines—GPU compute:
Azure NC-based instances are powered by NVIDIA Tesla K80 GPUs and provide the compute power required to accelerate the most demanding high-performance computing (HPC) and AI workloads. Customers can now use these instances to run deep learning training jobs, HPC simulations, rendering, real-time data analytics, DNA sequencing, and many more CUDA (Compute Unified Device Architecture)–accelerated tasks. Additionally, customers have the option to utilize RDMA (Remote Direct Memory Access) over InfiniBand for scaling jobs across multiple instances. Using InfiniBand between instances provides close to bare-metal performance when scaling out to tens, hundreds, or even thousands of GPUs across hundreds of nodes—allowing customers to submit tightly coupled jobs like Microsoft Cognitive Toolkit (CNTK)–based training for natural language processing, image recognition, and object detection.
Azure NV virtual machines—GPU visualization:
Azure NV-based instances are powered by NVIDIA Tesla M60 GPUs and provide NVIDIA GRID capabilities on demand. Scientists, designers, and engineers can now utilize these new instances for running hardware-accelerated workstation applications, designing the next concept car, or creating the next blockbuster movie. These instances support applications utilizing both DirectX and OpenGL.
Find the pricing details about these new N-series VMs here.