Microsoft has given a launch date for Azure N-Series. Corey Sanders Director of Compute, Azure says the virtual machines will move to general availability on December 1, 2016. The N-Series has been available in preview since August and the he VMs are split into two categories, NC and NV.
Azure N-Series virtual machines are powered by Nvidia GPUs and allow customers to access accelerated computing experiences. Microsoft says the VMs will launch in South Central US, East US, West Europe and South East Asia, all available on December 1st.
The company says the preview of the N-Series was a success, with thousands of customers participating. Microsoft has used customer feedback to enhance performance on the VMs ahead of a full roll out.
As mentioned, there are two types of Azure N-Series VMs:
Azure NC Virtual Machines
Azure NC VMs are powered by the Nvidia Tesla K80 GPU. Microsoft says there is sufficient compute power to deal with the most demanding HPC and AI workloads. Deep learning training jobs, rendering, real-time data analytics and DNA sequencing can be run in instances.
Customers can use Remote Direct Memory Access (RDMA) over InfiniBand to scale projects over numerous instances. The company explains how this spread workload runs on the GPU:
“InfiniBand provides close to bare-metal performance even when scaling out to 10s, 100s, or even 1,000s of GPUs across hundreds of machines. This will allow you to submit tightly coupled jobs using frameworks like the Microsoft Cognitive Toolkit (CNTK), Caffe, or TensorFlow, enabling training for natural language processing, image recognition, and object detection.”
Azure NV Virtual Machines
NV-based instance are also powered by Nvidia, this time the lower powered Tesla M60. This brings Nvidia GRID capabilities on demand. “Scientists, designers, and engineers can now utilize these new instances” to run workstation application. This could be for enhanced design experiences or editing tools.