Cooperation of Nvidia and Microsoft aimed at accelerating the application of artificial intelligence
The company Nvidia has told about cooperation with Microsoft, aimed at accelerating the application of artificial intelligence (AI) for the enterprise environment. The acceleration is achieved through the use of specialized Cognitive framework Microsoft to task an AI that uses the power of the GPU Nvidia that is available in the Microsoft Azure cloud or locally on the supercomputer Nvidia DGX-1.
Allegedly, Microsoft Cognitive teaches and evaluates algorithms for deep learning faster than other similar tools, being effective and scalable while preserving accuracy. Good scalability, in particular, means that the development can be tested on the local supercomputer Nvidia DGX-1 or another system with Nvidia GPU and then transfer them to the Azure cloud. This provides a significant performance gain in comparison with systems on the CPU. For example, on a supercomputer Nvidia DGX-1 framework runs 170 times faster than a server with two CPU Intel Xeon E5-2699v4 (44 cores).
While Azure virtual machines N on GPU Nvidia provided for testing only users with Azure, but soon they will be available to all users.