News
VMware Explore: VMware (and NVIDIA) Execs Promise To Deliver on Generative AI's Promise
VMware on Tuesday debuted a new solution, jointly developed with NVIDIA, that aims to address the legal, privacy and security concerns around the use of generative AI in enterprises.
VMware Private AI Foundation, expected to launch early next year, is an end-to-end platform designed to help organizations benefit from generative AI technologies while ensuring that their data remains private, secure and compliant. Powering it will be VMware's multicloud and NVIDIA's compute solutions.
VMware CEO Raghu Raghuram announced the Private AI Foundation during the opening keynote of the 2023 VMware Explore conference. He was joined onstage by NVIDIA CEO Jensen Huang, who touted the product as "reinventing enterprise computing."
[Click on image for larger view.] Private AI Foundation (source: VMware).
"For the very first time, enterprises around the world will be able to do private AI at scale ... and know that it's fully secure," Huang said.
With Private AI Foundation, businesses will be able to custom-create and train large language models (LLMs) on their proprietary data. Per NVIDIA's press release, the LLMs run "adjacent" to the customers' data, whether it's in the cloud, their own datacenters or the edge.
Raghuram stressed during the keynote that Private AI Foundation is key to helping organizations take advantage of the promise of generative AI, which he sees as threefold: the "general applicability" of LLMs, "universal access" that doesn't require data science expertise, and up to "100 times improvement in the economics of creativity."
Eventually, Raghuram predicted, generative AI will underpin the majority of modern application development. "Five or 10 years from now, calling something 'AI-powered' will be like calling something 'database-powered' now," he said.
For enterprises, however, the biggest barrier to taking full advantage of generative AI is concern for the security of their data. The concept of "private AI" navigates that barrier by taking what Raghuram described as an "architectural approach" to meet the specific privacy and compliance needs of an organization while using the cloud to facilitate access to LLMs and generative AI technologies.
"This is inherently a multicloud problem" that requires a multicloud solution, per Raghuram.
Other Announcements
Tuesday's VMware Explore opening keynote was a relatively quick one-hour affair, but it was not short on announcements. Besides the Private AI Foundation news, VMware also announced.
- vSAN Max, a new vSAN product that provides petabyte-scale, disaggregated storage.
- The vSphere+ lifecycle management service for ESXi.
- NSX+, a multicloud management, security and load balancing service.
- The ability to create virtual private clouds on demand using APIs.
- The option to perform automated, concurrent ransomware recovery across multiple VMs.
- The Tanzu Application Engine, designed to streamline requests from developers.
- Tanzu Intelligence Services, which use AI to automatically improve things application performance, resource management and more.
- Improvements to Apps on Demand.
- Improvements to Edge Cloud Orchestrator, née SASE Orchestrator.
Beyond these announcements, for many VMware stakeholders, the elephant in the room is the impending acquisition of VMware by Broadcom for a cool $61 billion. The acquisition was first announced over a year ago. At VMware Explore on Tuesday, Broadcom CEO Hock Tan appeared via a taped message to assure watchers that the acquisition is set to be completed, as scheduled, by the end of October 2023.
Stay tuned for more AI-related news from the conference, which concludes on Thursday. In the meantime, check out our other conference coverage, which includes: