Announcements and Final Recap — Virtualization Review

In Depth

VMware Explore 2024: Announcements and Final Recap

I had a great first day at VMware Explore 2024 despite my concerns about how Broadcom’s acquisition of VMware would affect it. The sessions that I attended were well-attended and full of technical content. The Expo Floor had around 50 vendors, and the hands-on labs (HOL) seemed popular. Overall, it was a bit quieter than in previous years, but it was enjoyed by me and the participants that I talked to.

In this article, I will discuss the rest of the event, emphasizing the keynote session, during which announcements were made regarding VMware’s future direction and products.

The Keynote
The keynote, which kicked off the event’s second day, centered around four major announcements: VMware Cloud Foundation 9, VMware Private AI Foundation with Nvidia, VMware Tanzu Platform 10.0, and Software-Defined Edge Innovations.

Broadcom’s CEO, Hock Tan, first took the stage and didn’t mince words while presenting his future of the company and IT in general.

Tan said that in the past, CEOs’ decisions to push their companies into the public cloud had “left their IT departments with post-traumatic stress disorder.” Silos in datacenters have left the IT department “screwed,” as many customers are now struggling with the three C’s of the public cloud: cost, complexity, and compliance.

He said the public cloud is more expensive than they expected and has an extra layer of complexity to manage. Complying with regulatory requirements is far more complex and costly than they anticipated.

Tan discussed VMware’s history and role in revolutionizing virtualization and cloud computing, but he acknowledged that much still needed to be done. He described VMware’s strategy as focusing on three key pillars: cloud, modern applications, and digital infrastructure.

VCF9
VMware Cloud Foundation 9 (VCF), which will be available soon, comprises VMware vSphere, vSAN, NSX, and vRealize. It will be more tightly integrated and managed, including a self-service cloud portal and a single console for operations and automation. Tan said it would be more secure and less cost than the public cloud. He also mentioned that all the packages would be simultaneously delivered as a single unit, so there would no longer be compatibility issues.

One of the more exciting features in VCF 9 is Advanced Memory Tiering with NVMe, which leverages less expensive PCIe-based Flash NVMe devices to act as a second layer of memory, increasing available memory within the ESXi host. This optimizes performance by directing VM memory allocations to NVMe devices or the host’s faster dynamic random-access memory (DRAM). This increases memory footprint and workload capacity while reducing the total cost of ownership (TCO).

VCF 9 will not just be available on-premises; it will also be available on the hyperscalers Microsoft Azure, Amazon Web Services (AWS), and Google Cloud Platform, as well as more than 300 managed services providers and OEMs like Dell, Hewlett Packard Enterprise, Lenovo, and Hitachi.

Private AI with NVIDIA
Private AI with NVIDIA was announced last year at VMware Explore 2023 in Las Vegas. It delivers a private and secure generative AI platform for enterprises. This year, VMware announced new capabilities for it.

The two major enhancements are model stores and guided deployments. Model Store allows MLOps teams and data scientists to provide secure LLMs with integrated role-based access control (RBAC) to ensure the governance and security of the environment and the privacy of enterprise data and IP. Guided Deployment allows workload domain creation workflow and deployment of other VMware Private AI Foundation components with NVIDIA to become significantly more streamlined. This can dramatically increase deployment speed.

We were also allowed a peek at some of the features that may be coming in the future. vGPU profile visibility will enable administrators to view all their vGPUs using vCenter. GPU reservations will allow admins to reserve resource pools for vGPUs. This will allow for better capacity planning, improving performance and operational efficiency. Data Indexing and Retrieval Service will enable you to chunk and index private data sources (eg, PDFs, CSVs, PPTs, Microsoft Office docs, internal web or wiki pages) and vectorize the data. This vectorized data will be made available through knowledge bases. AI Agent Builder Service will assist application developers, MLOps engineers, and data scientists in building and deploying AI Agents by utilizing LLMs from the Model Store and data from the Data Indexing & Retrieval service.

Finally, Broadcom will add new ISVs (Independent Software Vendors) and SIs (System Integrators) to the VMware Private AI Foundation with NVIDIA. This includes Codeium, HCL, Tabnine, and WWT.

Software on the Edge
Although I could not attend any session on Edge computing, VMware announced enhancements to its VMware SASE (Secure Access Service Edge) platform and introduced the Edge Compute Stack. VMware’s Edge Compute Stack will provide the infrastructure and tools necessary to support edge computing workloads, enabling the deployment and management of applications closer to where data is generated and consumed.

The key features of VMware’s Edge Compute Stack include a lightweight version of VMware’s virtualization technologies, unified management, resilient and secure operations, Kubernetes support, real-time data processing and analytics at the edge, and scalability from a few to thousands of edge locations.

Dance 10.0
Tanzu is a VMware product that can run and manage VM-based containers on public or private clouds with Cloud Foundry or Kubernetes (K8s).

Tanzu Platform 10.0 is its latest iteration and promises to deliver a self-managed, global control plane with greater automation and visibility across all application environments. It will have tighter integration with Spring. Tanzu AI Solutions is a set of tools for adding AI capabilities within apps. It will have Java-based APIs that will enable access to LLM models. An OpenAI-compatible API gateway can be used to broker hundreds of LLMs. It will have an observability toolst for testing and troubleshooting AI apps. Finally, it will add support for VMware Private AI Foundation with NVIDIA.

The Explore Party
As always, VMware held a party for its attendees to close out the event. Although there were lines for the food (sliders, hotdogs, etc.) at the party’s start, they were minimal to non-existent after the initial rush. They had a well-staffed bar with few lines, which the attendees appreciated.

On the floor were neon teeter-totters for the participants to teeter on. At the start of the evening, a DJ played a good mix of old and new hits, and the music wasn’t so loud that you couldn’t hold a conversation.

The headliner band, Jimmy Eats Worldcame on around 8:30. They played most of their old hits, including Pain, The Middle, Big Casino, and some lesser-known songs. The crowd was energetic and happy.

The food was filling, the drinks were plentiful, and it was a nice send-off to the event.

The Attendees
I enjoy talking to regular attendees. At lunch on the event’s last day, I was fortunate enough to sit at a table and have lunch with three first-time Explore attendees: one from upstate New York, one from Hawaii, and one from northern Washington. It was interesting to hear their perspective on the show. They all said that they attended to further their knowledge about VMware products and that they enjoyed the quality of the content presented. They were excited to participate in the event and amazed at the number of people and vendors there. I asked them about the pricing and licensing changes, and they all said they weren’t affected by them. I asked them about the events Private AI focus, and they all said it was interesting, but their companies weren’t at the point where they would be using it immediately. Two said they got to start thinking of AI projects, and if it were as simple to implement as VMware claimed, they would begin to work with it.