Nvidia and VMware today announced the introduction of an enterprise-grade hybrid cloud on AWS optimized for machine learning, AI, and data science workflows.
The VMware Cloud on AWS with Nvidia is capable of operating both from cloud and on-premise servers and will make it easier to migrate VMware vSphere-based applications to the cloud to accelerate high performance computing or machine learning for research, experimentation, or deployment in production.
The VMware Cloud on AWS will offer EC2 instances with GPU acceleration from Nvidia’s T4 100. AWS began to offer instances powered by Nvidia’s T4 100 GPU in March.
The news comes today at the beginning of the VMworld conference in San Francisco, and just days after VMware acquired cloud security and app development startups Pivotal and Carbon Black for $5 billion.
Nvidia also introduced vCompute Server today for the GPU accelerated deployment of workloads in virtual environments, including VMware’s vSphere, vCenter, and VMware Cloud. The latest virtual GPU offering from Nvidia can speed the completion of deep learning training up to 50 times faster than with a CPU alone. vCompute Server can also control GPU sharing for the distribution of compute power from a single GPU to multiple apps or virtual machines.
Users will be able to migrate vCompute Server workflows from on-premise data centers to the cloud with one click. Once in the cloud, needs can be modulated using AWS elastic infrastructure, growing or shrinking compute based on project needs.
“Together with VMware, we’re designing the most advanced and highest performing GPU accelerated hybrid cloud infrastructure to foster innovation across the enterprise,” Nvidia CEO Jensen Huang said in a statement shared with VentureBeat.