Nvidia’s AI Enterprise software solution offers new perspectives for researchers at the University of Pisa

It’s no secret that Nvidia Corp. has spent years developing a broad practice of artificial intelligence. A press release described the company’s work in the field as a “leading AI computing platform”.

In partnership with VMware Inc., Nvidia leveraged the vSphere cloud virtualization platform to deploy its AI Enterprise software suite in conjunction with Dell EMC VxRail and PowerScale for a robust enterprise solution.

“At VMworld last year in 2020, the CEOs of both companies came together and announced that we were going to bring our entire Nvidia AI platform to the enterprise on top of vSphere,” said John Fanelli ( photo, left), Vice President. of product, virtual GPU, at Nvidia. “The end impact is that AI is now accessible for the enterprise in the consumer data center. With Nvidia AI Enterprise and VMware vSphere, they can manage their AI the same way they are used to managing their data center today. There is no recycling, there are no separate clusters and there is no shadow IT. For the developer and the researcher, it makes things transparent.”

Fanelli spoke with furry jeans, host of SiliconANGLE Media’s theCUBE livestream video studio. He was joined by Maurizio Davini (pictured, right), Chief Technology Officer of the University of Pisa, and they discussed how the school leveraged Nvidia’s AI Enterprise solution for its researchers and the value of removing complexity from the process using predefined containers. (*Disclosure below.)

Benefits of virtualization

The tight integration of VMware and Nvidia software allows organizations to virtualize multiple technologies inside network systems. Organizations can share GPUs within servers, allowing multiple data scientists to accelerate deep learning workloads.

One of the users of this process is the University of Pisa, which houses a Dell Technologies and VMware Center of Excellence. The university’s IT group regularly tested new technology provided by the companies, including Nvidia’s artificial intelligence solution.

“We decided to integrate our virtual infrastructure with AI resources so that we could use it in a different and more flexible way,” Davini said. “We were able to show that the performances on the virtual and bare metal worlds were almost the same. In the virtual world, you are much more flexible, you can reconfigure everything faster and provide design solutions to researchers in a more flexible and efficient way. »

Nvidia and VMware’s work with the University of Pisa has also provided new insights into finding new ways to simplify the deployment and management of AI systems.

“Despite all the benefits and business that AI brings, AI can be quite complex,” Fanelli said. “We’re bringing in pre-made containers that take out some of the complexity. Containers allow you to do everything from initial data preparation and data science using things like Nvidia Rapids to training using solutions like PyTorch and TensorFlow. It helps this AI loop become accessible; this AI workflow is something a company can manage as part of its common core infrastructure. »

For the university, the simplified management of the AI ​​workflow has been a clear advantage. Virtualization makes it easier to migrate workloads from one data center to another, providing the kind of flexibility that researchers appreciate at the Italian school.

“The fact that the software stack has been simplified is something that has been very well accepted,” Davini noted. “The result of this work was very important for our research group.”

Watch the full video interview below and be sure to check out more CUBE conversations from SiliconANGLE and theCUBE. (* Disclosure: Dell Technologies Inc. sponsored this segment of theCUBE. Neither Dell nor other sponsors have editorial control over the content of theCUBE or SiliconANGLE.)

Photo: SiliconANGLE

Show your support for our mission by joining our Cube Club and our Cube Event community of experts. Join the community that includes Amazon Web Services and Amazon.com CEO Andy Jassy, ​​Dell Technologies Founder and CEO Michael Dell, Intel CEO Pat Gelsinger, and many other luminaries and experts.