Introduction
The VMware to AWS migration trend has reached a tipping point in 2026.
Following the Broadcom acquisition, enterprise CTOs are no longer just exploring the cloud; they are executing aggressive exit strategies to avoid a 300% to 1,500% increase in renewal quotes.
Every quarter, I watch more teams walk away from VMware. The pace picked up in 2024 and has only accelerated in 2026. As someone who has led four major migrations in the last eighteen months across manufacturing, healthcare, and retail, I’ve seen firsthand why AWS remains the primary destination for those fleeing the "Broadcom tax."
While Azure and Google Cloud often make the shortlist, AWS consistently wins on tooling maturity, Graviton-based cost savings, and the new agentic AI capabilities of AWS Transform.
Below are the seven primary reasons driving the rush for VMware to AWS migration in 2026.
Top 7 Reasons for VMware to AWS Migration
Here’s a quick overview of the seven key reasons why enterprises are rushing to migrate from VMware to AWS.
1. Perpetual License Model Replaced with Subscriptions
Before 2023, a VMware license was a one-time purchase you could work with for years. That went away in December 2023 when Broadcom moved everyone from perpetual licenses to a subscription-based model. The renewal quotes that followed have landed in the 800 to 1,500 percent range, higher than what these companies were paying before. Two of our clients called me the same week their 2025 numbers came through, both sitting in the upper half of that band. The comfort of distributing the costs of the one-time fee over multiple years was no more.
With AWS already running alongside VMware at most of these companies, AWS cost optimization suddenly looked a lot more attractive than a one-line VMware renewal quote. By that point, it was about which vendor they'd rather keep writing checks to, and AWS was the one that gave them more room to move.
2. The 72-Core Minimum Rewrote the Math for Smaller Servers
In April 2025, Broadcom changed VMware's licensing to require a minimum of 72 cores per server, regardless of the actual core count on the machine (Source). Previously, the minimum was 16 cores. This sudden change hit smaller-host environments in a bad way.
| Infrastructure | Core Licensing Model | Economic Impact |
|---|---|---|
| Legacy VMware | Licensed by actual physical cores (e.g., 16 or 24) | Predictable, proportional costs |
| VMware (Post-Broadcom) | 72-core minimum per subscription | Forced 3x cost increase for 24-core servers |
| AWS EC2 Native | Pay for exact instance size (e.g., m7g.large) | Zero waste; pay only for utilized vCPUs |
One of our clients had been running 20-core hosts across twelve sites for years, and they were paying VMware for the cores each server actually held. The 72-core minimum changed that. Every one of those hosts suddenly had to be licensed as a 72-core machine, and their renewal came in at almost three times what they paid the year before.
AWS handles this differently. You only pay for the exact instance size you pick, and the bill goes up or down based on what you actually use.
3. The SKU Catalog Collapsed
Broadcom reduced VMware's SKU count from about 168 to just four. Clients who'd been running specific vSphere plus vSAN combos for years can't buy those combos anymore. They're paying for NSX and a lot of other features that nobody's using.
One of my manufacturing clients, we did the licensing math last spring, and VMware was already the more expensive option compared to AWS. And that's just on the infra side. AWS keeps this simple. You pay for what you actually use. S3 for storage, EC2 for compute, and that's it.
4. AWS Transform Reduced the Migration Timeline
When we ran our first VMware to AWS migration in 2024, dependency mapping alone ate six months of my team's calendar.
AWS Transform launched in May 2025 with AI agents that handle dependency mapping, code refactoring, and database migration through to execution. On the migration we wrapped in Q1 2026, the same discovery work took nineteen days with AWS transform. I went in skeptical, but the numbers changed my mind. AWS has been investing in migration tools for two years now, and the time savings show up on every project calendar we run.
Usually, I recommend AWS migration services for clients to migrate from any infrastructure or environment to AWS, as trying to do it on their own usually leads to missed dependencies, unexpected downtime, and workloads that behave differently in AWS than they did on VMware.
5. Graviton Made Each Workload Cheaper to Run
The cost per workload changes the moment you leave VMware. On AWS, most of the workloads we migrate end up on Graviton, which is cheap enough that the monthly bill comes in lower than what the client was paying on VMware, hardware, and subscription combined.
VMware gives you one bill a year with no breakdown. AWS gives you a monthly bill with a clear breakdown for each workload. If something costs more than it's worth, you can just shut it off and see the savings right away.
6. AWS Bedrock Brought AI Into the Same Bill
VMware doesn't have any strong AI features. Adding generative AI to a workload on VMware means a separate ML pipeline, a new vendor contract, and a fresh security review for every model the team wants to use.
On AWS, the same workload has Anthropic, Meta, and Mistral on one API, just a call away through Bedrock. The data stays in the VPC, and the AI spend shows up on the invoice that the workload already sits on.
For CTOs who were working with VMware, they had to explain to their boards why AI adoption is not feasible at the moment, and how the whole setup will take months to get in place, and still be hard to maintain. But, after opting for VMware migration to AWS, they can easily adopt AI for their workloads and stay ahead of the trend.
7. The Talent Market Has Already Shifted
When we staff a migration project, the engineers we bring in have almost no VMware background anymore. Out of the last eleven we hired for infrastructure work, one had worked on vSphere. All eleven had worked on AWS. After Broadcom took over VMware and changed the licensing, most engineers saw where things were headed and moved their skills to AWS or Azure. The ones entering the market now never learned VMware in the first place, because no one was hiring for it by the time they started their careers.
Hiring for an AWS team may take weeks. But hiring for a VMware team can take up months or even quarters, and every year, there are fewer people to hire.
Final Note
Every VMware to AWS migration project I've worked on in the last two years has started with one of the seven reasons above. Most end up touching four or five before cutover. The licensing pressure is what gets the project approved. The tooling, the Graviton savings, and the talent market are what make it work once it starts.
The migration window is short, but it carries the most risk. A missed dependency, a bad cutover weekend, or a wrong call on post-migration costs can undo the reason why the CFO chose to migrate.
And, if your team is planning a move and doesn't have engineers who have run one before, you can hire AWS developers for the migration window and move operations back in-house once the workloads are running as expected.
Frequently Asked Questions (FAQs)
1. Is there a minimum core requirement for VMware in 2026?
Yes, Broadcom has enforced a 72-core minimum per subscription for most new enterprise agreements, which can significantly inflate costs for smaller servers.
2. What is AWS Transform for VMware?
AWS Transform is an AI-powered migration service that uses agentic AI to automate discovery, dependency mapping, and the generation of migration plans for VMware workloads.
3. Can I opt for VMware to AWS migration without refactoring my apps?
Yes, VMware Cloud on AWS (VMC) supports lift-and-shift with zero downtime, and services like AWS MGN automate the server replication side of it. Most long-term savings, though, come from moving to native AWS EC2.
