If a picture is worth a thousand words, then VMware Inc.’s current narrative was written in a three-minute video that introduced the company’s keynote session during its Explore conference in Las Vegas today.

The video featured brief remarks from several top enterprise company executives on the subject of generative AI, including Amazon Web Services Inc. Chief Executive Adam Selipsky, Google Cloud CEO Thomas Kurian Advanced Micro Devices Inc. CEO Lisa Su and Hock Tan, CEO of Broadcom Inc. The presence of these leading executives captured VMware’s interest in simplifying its clients’ use of multiple clouds, building new alliances with chipmakers, and preparing for its imminent acquisition by Broadcom.

This week’s VMware Explore gathering comes at a time when the use of generative AI is attracting widespread enterprise interest. The company made several announcements today in this area, including VMware Private AI Foundation with Nvidia, a platform with generative AI software for building, training and deploying models out of the box.

“Our goal is to dramatically decrease the complexity for what it takes for an enterprise to be productive with AI,” VMware CEO Raghu Raghuram (pictured, right, with Nvidia CEO Jensen Huang) said during a briefing for the media on Tuesday following his keynote remarks. “We are providing customers with their choice.”

VMware’s expanded partnership with Nvidia Corp. highlights enterprise concern around the risks that come from using proprietary business data to train generative AI large language models in applications. Private AI Foundation offers enterprise customers the ability to run AI services adjacent to data with the goal of preserving data privacy.

The announcement also provides developers with access to Nvidia’s AI Workbench to test and customize AI models, and the platform includes NeMo, Nvidia’s cloud-native framework for deploying generative AI models nearly anywhere.

“This is groundbreaking computer science,” said Huang, who appeared on the Explore keynote stage with Raghuram. “For the first time, enterprises around the world will be able to do Private AI at scale and deploy it through your company.”

‘Cloud smart’ approach

VMware’s generative AI announcements were coupled with a slew of new releases for multicloud support. Generative AI and multicloud are intertwined in the company’s view and this will drive customer decision-making.

The challenge will be developing AI architectures that are not based on a single cloud platform, but rather can be deployed as a consistent, secure and cost-efficient operating model across various clouds. This will require a “cloud smart” approach, according to Kit Colbert, VMware’s chief technology officer, who expressed an interest in avoiding previous sole-source cloud limitations.

“How are we going to architect these AI systems and are we going to learn from the past?” Colbert asked during the post-keynote press briefing. “This is the opportunity we have by taking a cloud smart approach with AI. This is going to be a multicloud problem.”

There was no shortage of multicloud announcements from VMware today. The company kicked off its conference with numerous enhancements including extension of its NSX multicloud software called NSX+ to support consistent networking and security across VMware Cloud environments. VMware also announced the integration of AI features into its Anywhere Workspace cross-cloud services portfolio and new orchestration capabilities for the edge, the next multicloud frontier.

A key element of VMware’s multicloud message is that its partnerships with hyperscale cloud providers allow customers to run cloud solutions directly on its virtualized infrastructure.

“We run our stack on metal from these partners,” Krish Prasad, senior vice president and general manager of VMware’s Cloud Infrastructure Business Group, said in an exclusive interview with SiliconANGLE. “There is no intermediate layer. That’s the beauty of this.”

Acquisition gets closer

In remarks today, VMware executives remained cautiously optimistic that the company’s acquisition by Broadcom would finally come to a long-awaited conclusion. Movement toward final approval gained steam earlier this week when regulators in the United Kingdom greenlighted the $69 billion deal, although additional clearances are still needed by numerous countries, including China.

Asked during the press briefing about whether he was concerned about China’s inaction to date, Raghuram sidestepped the question and simply said, “We are well underway with the regulatory process with the remaining geographies.”

Broadcom’s Tan offered little in his own remarks during a short video aired during the keynote, other than to express confidence that the deal would close during Broadcom’s fiscal 2023 ending Oct. 30 and that the company is committed to its investment in VMware. “We will focus on investing in VMware products and services,” Tan said.

Photo: VMware/livestream