AWS becomes the first cloud services in launching new compute instances, beating Google and Microsoft on the way. The new instances aim at allowing developers for building apps – the apps that can derive Nvidia’s new gen Volta GPUs. These GPUs have the designs for providing high-performance acceleration for applications such as AI computation.
One of the newest necessary trends of today among the firms is hinging on to machine learning for enhancing business aspects. To be successful in the same, building new models need vast computation attributes. This is where Volta comes in as the leader in the industry, and it even beats the past generations of silicon of Nvidia. Amazon cloud has made it available for the businesses to grab hold and make more out of the future possibilities.
Users can now run the instances with up to 8 V100 GPUs. Amazon’s data centers in Ireland, Tokyo, Oregon, and Northern Virginia will initially roll out these instances.
Furthermore, Nvidia came up with the launch of a new GPU Cloud that aims to serve companies providing ‘most optimized environment for running deep learning applications’ – using Nvidia’s hardware in the public cloud. Developers can sign up for their respective GPU Cloud accounts for free of cost. They can further connect their credentials to an Amazon Machine Image (AMI) – created by Nvidia for machine learning.
The AMI can offer developers an opening to deploy containers on AWS. These containers are optimized for usage of various machine learning frameworks with CUDA – a library created by Nvidia for processing the machine learning calculations with its hardware.
The GPU Cloud has compatibility with both new Volta GPUs and old Pascal hardware series.
Nvidia has its future plan to offer this GPU Cloud software to other providers as well. Although, Nvidia is yet to have a timeline for this action. Apparently, other providers are going to follow the suit in the coming future, as Amazon launched the Volta hardware on its own cloud.