The Hidden Costs of VMware: Why Enterprises Are Paying Billions More

Earlier this year, T-Mobile sued Broadcom over VMware support for tens of thousands of virtual machines running across more than 300,000 CPU cores. The number is striking, but the more important detail is that T-Mobile is not an isolated case. It is the third major enterprise in two years to end up in litigation over the same issue: what happens when a company builds its entire operational stack on top of a single vendor, and that vendor changes the terms.

For investors focused on capital markets and infrastructure, this is not a story about software licensing. It is a case study in what vendor lock-in actually costs, and why compute sovereignty is quickly becoming a capital allocation question rather than a technical one.

What Actually Happened

Broadcom closed its $69 billion acquisition of VMware in November 2023. Within months, the company eliminated perpetual licensing entirely, moving every customer to subscription pricing, and consolidated products into a small number of expensive bundles regardless of what individual customers actually needed.

In April 2025, Broadcom raised the minimum licensable deployment from 16 cores to 72 cores, instantly tripling the cost floor for smaller deployments that had no intention of running anything close to that scale.

The price increases that followed are documented, not anecdotal. European customers have reported increases as high as 1,500%. AT&T's costs were set to rise 1,050% under the new terms, a figure confirmed in the company's own court filings before the two sides settled privately in 2024. Industry-wide reports cite increases in the 150% to 300% range as the norm, not the exception.

The litigation followed. AT&T sued and settled. Tesco is currently in the process of migrating 40,000 server workloads off VMware while accusing Broadcom of abusive conduct. T-Mobile offered $5.28 million to extend support for one additional year and was refused, then offered $20 million for two years and was refused again. A judge ultimately granted T-Mobile a temporary injunction to keep support running through August 2026 while the case proceeds.

Meanwhile, Broadcom's VMware revenue grew 13% year over year in the first quarter of fiscal 2026, with total contract value exceeding $9.2 billion. The lawsuits are real. The customer backlash is real. The revenue growth is also real, at the same time. That is the entire story in one sentence: Broadcom is generating more revenue by extracting more value from customers who cannot easily leave.

Why Migration Is Not a Simple Answer

The obvious response is that affected companies should move to alternatives like Proxmox, Nutanix, or Hyper-V, all of which have gained real traction since 2024. Many are trying. But T-Mobile's own legal filings explain why this is not a straightforward project at enterprise scale: more than 1,000 applications, tens of thousands of virtual machines, a network that cannot tolerate downtime, and a multi-year technical migration that has to happen while the business continues operating without interruption.

Broadcom understands this dynamic well. Every price increase and every refused support renewal is priced against exactly how expensive and slow it is for a company of that size to actually exit the relationship. This is not a licensing dispute. It is a company discovering it built its entire infrastructure on a vendor's continued goodwill, and finding out that goodwill expired the day ownership changed hands.

The Part That Should Concern Every Infrastructure Investor

Much of the current conversation around vendor concentration focuses on hyperscale cloud providers and AI compute specifically. VMware is the same failure mode wearing a more familiar, less discussed uniform, and that is precisely why it deserves attention. The underlying risk is not unique to AI infrastructure. It is inherent to any critical system built entirely on a single vendor's continued cooperation.

Broadcom did not break any rules. It identified the protections nobody had bothered to negotiate, because for two decades VMware was stable, reasonably priced, and not owned by a company with a documented history of acquiring mature software assets and extracting maximum value from captive customer bases. That approach has worked before, and it is working again. T-Mobile, AT&T, and Tesco are large enough and motivated enough to fight it in court. Every smaller company running the same infrastructure, without the legal budget to take on a company of Broadcom's scale, simply absorbed the increase and moved forward.

The lesson here is not to avoid any particular vendor. It is that infrastructure decisions made when a provider is stable, well regarded, and reasonably priced need to be reevaluated the moment ownership changes, because the terms a company signed up for were never contractually guaranteed indefinitely. They were convenient for as long as the previous owner needed them to be. Control over one's own compute and data infrastructure is no longer a theoretical resilience exercise. It is the difference between being a customer and being a captive counterparty, and a growing number of large enterprises are finding out which one they actually were.

The Eliakim Capital Perspective

At Eliakim Capital, this story reinforces a thesis:

Infrastructure ownership and financing structure matter as much as the technology itself. Companies that finance and build compute capacity through structures that preserve operational control, rather than renting critical infrastructure

entirely at a single vendor's discretion, are structurally better positioned when ownership or pricing power shifts upstream.

The VMware situation is a preview of a dynamic we expect to see repeat across the compute and data center financing stack. As capital continues flowing into AI infrastructure, HPC deployment, and digital power capacity, the investors and operators who structure their capital and vendor relationships with exit optionality in mind will be the ones who are not forced to negotiate from a position of dependency when the terms inevitably change. Compute sovereignty is not a defensive posture. It is becoming a core underwriting criterion.

This article is provided for informational and educational purposes only and does not constitute investment advice. References to companies and litigation are based on publicly reported information and are intended solely for analytical discussion.