By Broadcom’s 2023 acquisition of VMware triggered one of the most disruptive vendor relationship collapses in enterprise IT history. Overnight, decades of trust between VMware and its customer base evaporated — replaced by forced licensing overhauls, elimination of perpetual licenses, and pricing restructured to extract maximum revenue from a captive installed base. For enterprises that had built their entire infrastructure on VMware’s hypervisor stack for 20-plus years, the message was unambiguous: your loyalty means nothing now.
The financial exposure is significant. VMware’s new subscription bundles force customers into the highest-tier plans regardless of feature utilization, with industry analysts estimating price increases of 3x to 10x for many organizations. But the deeper wound is cultural. Enterprise IT teams don’t just buy software — they build institutional knowledge, operational processes, and vendor relationships around platforms. When that relationship is severed unilaterally, the damage extends far beyond the budget line.
Consulting firms including Gartner estimate that approximately half of VMware’s customer base is actively evaluating or executing exit strategies. That represents one of the largest voluntary platform migrations in enterprise infrastructure history — and it is happening right now, simultaneously across thousands of organizations in the US, Europe, and beyond. The problem is that migration is complex, risky, and rarely as clean as a vendor promises.
The open source virtualization ecosystem has long existed as an alternative to VMware’s proprietary stack, but for most enterprises it remained a curiosity rather than a credible replacement. Operational complexity, support concerns, and the absence of enterprise-grade tooling kept most IT leaders on the VMware path even when they questioned the value. What changed is not just the pricing shock from Broadcom — it is the emergence of fully integrated, turnkey open source virtualization platforms built specifically to absorb VMware workloads at enterprise scale.
Vates, a French open source software company founded over a decade ago, has positioned its XCP-ng hypervisor and Xen Orchestra orchestration platform as the architecturally closest open source equivalent to the ESXi and vCenter stack. With over 500 enterprise customers running XCP-ng in production and the United States now representing its largest and fastest-growing market, Vates is at the center of the VMware displacement wave — and its CEO has a clear-eyed view of where this market is heading.
The Guest: Olivier Lambert, CEO and Co-Founder at Vates
Key Takeaways
- Approximately half of VMware’s customer base is actively seeking to exit, driven primarily by broken trust rather than pricing alone — and the inbound migration pipeline shows no signs of decelerating.
- XCP-ng mirrors the ESXi architecture and Xen Orchestra mirrors vCenter, making Vates the closest structural equivalent to VMware’s stack in the open source ecosystem.
- Vates prices per physical host — not per socket or memory — a fundamentally different model that eliminates the VMware licensing penalty for high-density or memory-intensive workloads.
- The standard enterprise migration pattern is phased: raise VM density on VMware to free hardware, run an XCP-ng POC on liberated hosts, migrate least-critical workloads first, and expand incrementally while VMware remains available as a fallback.
- EU tech sovereignty initiatives and US infrastructure resilience concerns are converging to accelerate open source virtualization adoption from two different geopolitical directions simultaneously.
***
👇 Click to Read Full Transcript & Technical Deep Dive
In this exclusive interview with Swapnil Bhartiya at TFiR, Olivier Lambert, CEO and Co-Founder at Vates, discusses the enterprise VMware exodus triggered by Broadcom’s acquisition, the architectural positioning of XCP-ng and Xen Orchestra as direct VMware alternatives, practical phased migration strategies used by hundreds of enterprise customers, the commercial sustainability model of fully open source enterprise software, and the geopolitical tailwinds driving tech sovereignty adoption across the US and European markets.
The Broadcom Acquisition and the VMware Trust Collapse
The Broadcom-VMware acquisition set off an immediate signal cascade in the enterprise IT market. Olivier Lambert traces Vates’ first awareness of the market shift to 2022, before the deal closed, and describes the inbound interest from enterprises as a sustained, uninterrupted flow that has not decelerated through 2025. The driver, he argues, is not simply pricing — it is a fundamental breakdown of the vendor-customer relationship that enterprises spent two decades building with VMware.
Q: Walk us through what you are actually hearing from enterprises right now. What is driving them to finally make a move?
Olivier Lambert: “The first signal we started to hear was back in 2022 when Broadcom made the first announcement, but it wasn’t official until 2023. We started to hear, as a software virtualization company, many people started to come to us. What’s funny is we started to hear noise from the US first. I think it’s because the US market is very well aware about who Broadcom is. Since then, in the last three years, it’s like a constant flow of people trying to seek to get out of VMware because of the pricing, which is the main driver to seek for an exit. But as you mentioned in the introduction, it’s a question of trust — a broken trust with them more than a pricing question. It was brutal. And now we see that around half of those VMware customers are seeking to get out of VMware in the end.”
Clearing the Open Source Misconception for VMware Refugees
Enterprise buyers coming from a proprietary VMware environment carry specific assumptions about open source software — assumptions that differ sharply from those of the existing open source community. Lambert describes two distinct audiences Vates serves and explains why, perhaps surprisingly, open source skepticism is rarely the primary barrier for VMware refugees evaluating XCP-ng. The questions enterprises ask first reveal what actually blocks migration decisions.
Q: What are the biggest misconceptions that you have to clear up before a real conversation can even start when you talk about alternatives?
Olivier Lambert: “It’s kind of very different than even 10 years ago. I would say we have two pillars right now. We have the open source community — users that are playing with machines in their garage and so on. We’re not trying to sell anything to them. They are ambassadors. They are very well aware of what open source is. And the other pillar is people coming from VMware. They are not really used to open source, but frankly, when they come to us, the open source question is maybe the third or fourth question. The first one is the cost and the expertise on the migration. They don’t really care about the fundamentals of open source. They know we are controlling the way we build the platform, we produce the security reports and so on. And for them it’s good enough — or even a security advantage against a potential trust problem that you had with proprietary software. Because even if, for whatever reason, the company disappeared or was acquired, the existing code has to stay there. So frankly, I was kind of surprised in the right way that open source is not an obstacle on the decision to migrate out of VMware. It’s more like: are you able to help us make the migration? Is your product working? Is it turnkey? I think that’s maybe the main concern about open source — because you could have very complex open source that you need to build everything yourself. But if you build a turnkey product, they don’t really care about that. It’s not a problem, at least.”
XCP-ng vs. ESXi and Xen Orchestra vs. vCenter: Architectural Positioning
Vates’ competitive differentiation starts with architectural familiarity. XCP-ng’s core engine is the Xen hypervisor — a project roughly the same age as ESXi — giving the platform a structural parallel to VMware’s stack that no other open source alternative can claim as directly. Lambert frames the feature parity debate honestly, acknowledging VMware’s R&D depth, while arguing that the majority of VMware customers were never using the features that made VMware unique — and were paying for all of them anyway.
Q: How does XCP-ng and Xen Orchestra stand out from competitors?
Olivier Lambert: “The core engine of XCP-ng is Xen, and it’s almost the same age as ESXi. In a way, those two platforms — the VMware one and the Vates one — are kind of similar in terms of architecture. Instead of ESXi, from one side you have XCP-ng, and instead of vCenter you have Xen Orchestra. Nobody has 100% of all the features of VMware, because it’s the best virtualization product feature-wise, with the most advanced R&D — there’s no question about this. But the thing is, most customers — more than 50% — aren’t using all the bells and whistles of VMware, and they are forced to get the highest plan, the one with absolutely everything they don’t need. So they’re seeking to do something else. The main differentiator of our platform is the fact it’s bundled into one solution. You don’t have to get the backup solution from outside — you can use our backup solution if you want to back up your XCP-ng VMs. If you want a multi-hypervisor backup, you’ll go to Veeam and so on. But for the essential stuff, you can have everything in one turnkey solution. The fact that we are both open source and — unlike classical open source — pretty much turnkey and scalable for even hundreds of hosts or a few thousand hosts, makes sense for them. And obviously pricing is a big differentiator: we price per physical host, not per socket or per memory. All those differentiators matter. And to be fair, we hire people coming from VMware because they’ve been leaving the company since the Broadcom acquisition. VMware was a very nice company before, especially for all the employees. Now with Broadcom, it’s completely different. So some of them are coming to Vates, and they are really helpful — sometimes VMware users even find back the technical account manager they had with VMware but now find them at Vates. It’s kind of like you never leave the family. Because as I said, trust is probably the biggest currency right now in that market. They want to trust a company, they want to trust a product, and this is what we are doing. I’m an engineer, so it hurts a bit to say this — but the product is maybe not the main reason people are migrating to our solution. It’s everything around it: the capability to help them make the migration. That’s a very important point.”
Enterprise Migration Methodology: Phased, Pragmatic, and Risk-Managed
For enterprise IT teams that have operated VMware infrastructure for two decades, migration represents a dual risk: the financial risk of remaining on VMware under Broadcom’s pricing regime, and the operational risk of moving critical workloads to an unfamiliar platform. Lambert describes in detail the phased migration pattern that Vates has refined across hundreds of enterprise deployments — a methodology designed to manage both risk vectors simultaneously without requiring new hardware investment upfront.
Q: What should enterprises realistically expect when they make the move away from a traditional proprietary virtualization platform? And how does Vates make that less painful?
Olivier Lambert: “I think the best way to answer it is to give you an example of how we’ve already done hundreds of migrations. The average enterprise customer is juggling between two main risks: the financial risk of staying at VMware, and the technological risk of leaving VMware for something else — because those people have been using VMware for the last 20 years. They are very used to the product and they don’t want to relearn something from scratch. So the product has to work easily as a turnkey product. The way migration is done is usually they don’t even repurchase new hardware. They migrate VMs onto a smaller number of physical hosts — raising the density on the VMware side — which frees up some hosts to reinstall something else. They reuse that hardware, now free from VMware, to install our stack. They test it, do a POC on top of it, and when they’re happy with the POC, they start migrating the less critical virtual machines to XCP-ng as their new production environment. Then it grows — slowly — from maybe 6, 8, 10 machines to 15, 20, 30, 50, 100. Sometimes replacing everything, sometimes not. When you have a very critical part of your infrastructure that you’ve trusted 24×7 for the last 20 years with VMware, you can keep maybe 10% of your infrastructure on VMware. You’ve solved the financial risk and also mitigated the technological risk. That’s how we enter large enterprise customers — first by getting a small part of the infrastructure, then growing as they move from less critical to more and more critical parts of the stack. We do have a V2V migration tool — VMware to Vates — that connects to the API of your ESXi or vCenter. You tick the boxes of the VMs you want to migrate, then you migrate them. You don’t remove the original VMs — they stay shut down on the VMware side — and for a few weeks you watch for results and see if your users have any complaints. If it’s fine, you move forward with the rest of the infrastructure. It’s a very pragmatic approach: less critical to more critical, then decide whether to replace the VMware infrastructure entirely or not.”
Market Trajectory: 500+ Enterprise Customers and the Peak of the Migration Wave
With over 500 enterprise production deployments in the US alone, Vates has accumulated enough market signal to characterize where the VMware exodus wave currently sits — and how much runway remains. Lambert’s assessment draws on Gartner’s projection that roughly half of VMware’s installed base will ultimately leave, combined with Vates’ own inbound pipeline data showing no deceleration as of the interview date. He also differentiates market behavior between the more aggressive US buyer and the more conservative European enterprise.
Q: You have over 500 enterprise customers running this in production in the US already. What does that base tell you about where the market is heading in the next two to three years?
Olivier Lambert: “It’s not the end of the wave right now. I would say maybe we’re closing to the top of it, but many consulting firms like Gartner are expecting half of those customers to leave. What’s really impressive — and it’s not just within the US even if the US market is the first one we’re selling in across all countries in the world — is that the US market is the fastest to move. It’s also probably the one that is more keen to take risks, to get out of VMware faster than maybe more conservative markets in Europe, for example. But still there is a constant inbound of people coming to us every day asking how to escape VMware. So I think it’s not even decelerating yet. Maybe we are close to the peak. But I think in one or two years we’ll reach the peak of active migration projects. Within Europe it might be in the next three to five years, because again it’s a bit slower to keep the pace there. But that’s the ballpark numbers we’re seeing right now.”
Tech Sovereignty: Geopolitical Tailwinds from Both the US and the EU
The VMware displacement story is increasingly intersecting with a parallel macro trend: the growing political and regulatory pressure on enterprises and governments to reduce dependency on single foreign technology vendors. In the US, this surfaces as infrastructure resilience thinking — the realization that IT sourcing carries strategic risk similar to supply chain sourcing in physical industries. In Europe, it manifests as a formal tech sovereignty agenda that actively favors European software vendors. As a French company, Vates sits at the intersection of both currents.
Q: A lot of countries are looking at tech sovereignty, AI sovereignty, data sovereignty, and sovereign clouds. What does that mean for Vates?
Olivier Lambert: “For the US markets, it’s more seen from the resilience angle — making sure, as an IT person within a large corporation, that you’re not relying only on one player. Which was the case before the Broadcom acquisition. I think people realize that sourcing your IT software is in fact like sourcing steel or anything in the regular industry. It was a reminder that sourcing is critical and you need to be able to not rely only on one player. That’s why some of them want to keep VMware, some are going to two different companies to source their virtualization platform, and some are using another Kubernetes platform while keeping the virtualization platform — just to make sure they’re not relying only on one player. For the European part of the world, there’s another angle: reducing dependency on extra-European companies. That’s why we’re seeing a pretty fast acceleration on the European market right now. If now 60% of our business is within the US, we think that in three or four years it will be more balanced with the European Union — which is, as I said, a bit slower to get started. But when they decide to focus on something, and given that we are a French company, that makes things very exciting for potential growth within Europe.”
Open Source Commercial Sustainability: Selling Expertise, Not Code
The commercial viability of fully open source enterprise software — where every line of code is public and freely copyable — is one of the most persistent questions in the industry. Vates operates on a model that Lambert describes as structurally advantaged by its target customer: VMware refugees who have spent 20 years paying for enterprise support and have no expectation of running critical infrastructure without a commercial support relationship. The result is a business model where open source and commercial sustainability are not in tension but are mutually reinforcing.
Q: How does Vates stay commercially sustainable while also keeping as much code as possible in open source?
Olivier Lambert: “I think we are kind of lucky at Vates for many reasons. If you are a customer coming from VMware, there is no way you won’t pay anything to be supported for your critical infrastructure. You’ve been paying VMware for the last 20 years, and it’s normal to have support if you have a problem. We are not addressing the existing open source community — they won’t pay anything, and that’s fine. That’s not where we want to survive. It would be impossible to survive if we were only targeting them. The main target we have are VMware customers, and for them it’s not even something that could be thought about — not paying anything and having the responsibility internally to maintain and solve bugs. Even if we are fully open source — you can download it, there is no restriction, it’s not even open core, it’s fully open — what we are selling is the expertise on the code we are producing. Not the code itself, because you can copy it, use it, whatever. But as it’s a very critical part of your infrastructure — not a SaaS calendar application or whatnot, it is running your entire infrastructure — it’s too critical to not have support, especially if you come from VMware where you had this kind of support. The open source business is not a commercial problem for us. We have the best of both worlds: we can do everything in open source the way we want, and we sell the expertise on the code we built. It’s a very sane model. We are fully aligned with our customers: they want to make sure everything works well, because if it does, we have fewer support tickets. Everyone is incentivized in the right way — they pay us to improve the software, and we improve the software to reduce the number of problems. We are fully aligned with our customers, even while being fully open source.”
