Nvidia Corp. is enabling artificial intelligence and machine learning in the public cloud with the launch of new “accelerated GPU services” on VMware Cloud on AWS.

Announced at the VMworld 2019 conference today, the new services make it simple to migrate existing vSphere-based applications and software containers to VMware Cloud on AWS, which is a hybrid cloud platform that runs VMware’s software-defined data center stack on Amazon Web Services Inc.’s public cloud.

Once those apps have been redeployed, they can then leverage Nvidia’s high-powered graphics processing units to take advantage of technologies, including high-performance computing, machine learning, data analytics and video processing applications, the company said.

Nvidia argues that AI workloads such as image and voice recognition, financial modeling and natural language processing are best done on its GPU hardware, as it significantly speeds up training and inference times when compared to traditional central processing units.

The new services leverage Amazon’s EC2 bare-metal instances and Nvidia’s new Virtual Compute Server software, and combine these with Nvidia T4 GPUs to accelerate AI workloads.

“From operational intelligence to artificial intelligence, businesses rely on GPU-accelerated computing to make fast, accurate predictions that directly impact their bottom line,” said Nvidia founder and Chief Executive Officer Jensen Huang. “Together with VMware, we’re designing the most advanced and highest performing GPU-accelerated hybrid cloud infrastructure to foster innovation across the enterprise.”

One of the biggest benefits of running AI workloads on VMware on AWS Cloud is they’ll be able to take advantage of the “elastic” nature of Amazon’s cloud infrastructure, Nvidia said. The new services will enable customers to scale their AI workloads as necessary, growing and shrinking their training environments depending on the needs of their data scientists.

Other benefits include improved security and manageability of AI applications, not too mention better portability. Customers will be able to migrate apps between VMware on AWS Cloud and on-premises environments with just a click of a button, with no downtime, Nvidia said.

Analyst James Kobielus of SiliconANGLE sister market research firm Wikibon said the partnership means VMware can now offer high-performance AI compute infrastructure to its customers, which it was unable to do before.

“With this partnership, VMware customers can easily move AI workloads that had been running on bare metal CPUs over to virtualized clusters of Nvidia’s latest and fastest GPUs running on VMware Cloud on AWS,” Kobielus said. “This should give VMware a leg up on other hybrid cloud solution providers such as IBM and HPE in the battle to run enterprise customers’ most advanced AI apps.”

Image: Nvidia/Vmware