AWS EVS: A Step Toward VCF Everywhere

Broadcom has ended sales of VMC on AWS directly from Amazon. It’s even further away from being an AWS-native solution. This isn’t an “end of life” announcement — customers already on VMC will still be supported, and you may buy it under less agreeable terms from Broadcom directly. Still, it is the end of the sales pipeline for AWS’s flagship VMware-managed cloud offering.

Why? Because Broadcom is unifying around VMware Cloud Foundation (VCF) as the standard VMware stack across all hyperscalers and private cloud. The goal is simple: eliminate the variability between hyperscaler offerings and put VMware’s own control plane at the center — whether the hardware belongs to AWS, Google, Oracle, Azure, or your own data center.

For years, AWS held the “prime position” as VMware’s go-to hyperscaler partner. That position is gone. EVS is AWS’s counterpunch:

“If VMware won’t sell a managed VMware-on-AWS service anymore, we’ll offer a self-managed VMware stack on our bare metal. Customers stay in AWS, workloads stay in AWS, and we don’t cede them to Oracle, Google, or Azure.”

The Three EVS Personas

EVS isn’t designed for the general market. It’s for a very specific, often “cornered” set of VMware customers:

1. The Data Center Exit Deadline
   Companies with a lease ending, M&A-driven consolidation, or corporate cloud-first mandate. They have hundreds of VMware workloads too critical or complex to re-architect in time. EVS offers a fast lift-and-shift that stays in AWS.

2. The VMC Customer in Limbo
   Customers happy with the managed VMC service but now facing Broadcom’s licensing and support changes. Their options: deal directly with Broadcom, jump to another hyperscaler’s VCF offering (OCVS, GCVE, AVS), or stay in AWS and manage VCF themselves.

3. The Hybrid-Cloud Architect
   Teams that value operational consistency across on-prem and cloud more than “managed” services. For them, running VCF themselves in AWS is alignment — not a burden.

AWS’s Checkbox Strategy

In product strategy terms, EVS is AWS “checking the box” — but it’s an important box:

• Retain workloads & market share – Without EVS, many of these VCF workloads would migrate to Oracle’s OCVS, Google’s GCVE, or Azure’s AVS.

• Complete the migration menu – EVS adds a “keep-as-is (on AWS)” option alongside re-host, re-platform, and re-factor.

• Future-proof spend – Once VMware workloads are inside an AWS VPC, AWS can upsell and cross-sell into native services over time.

The VxRail Parallel

If this decision feels familiar, it’s because Dell VxRail customers are in a similar position.

Just like VMC on AWS, you can’t easily buy new VxRail anymore. Current customers can continue running their existing hardware and software under support agreements, but new systems require customers to sign an affidavit. Dell executives have informed me that VxRail isn’t long for this world.

In both cases, customers are faced with a shift away from a vendor-managed comfort zone (VxRail integration/support or VMC automation) toward a future where the operational burden will change hands. And in both cases, the workloads involved — ERP, SAP, HR — are “always-on” systems of record that don’t easily move to new platforms.

The point is: whether it’s VxRail or VMC on AWS, the status quo has an expiration date, and the operational responsibility will end up somewhere new.

The Inertia Factor

A former VMC on AWS SRE put it bluntly:

“EVS eliminates a couple of the VMC pain points, but at the cost of supportability. AWS hardware fails at a prodigious rate — 2–5% per month. Without VMC automation to make that painless, customers will have a ton of pain points. Miss a maintenance step and you can lose your vSAN data.”

In the old VMC model, VMware’s automation masked these failures — workloads rebalanced automatically, failed hosts swapped, data preserved. In EVS, you are now the automation layer.

What IT Leaders Tell Me

When I talk to CIOs and IT decision makers, operational resiliency comes up every time. Reliability is not a “feature” — it’s the baseline for even considering a platform.

If a solution is perceived as less reliable than their current VMware environment, the conversation ends. No amount of “more control” or “lower cost” can compensate.

This is why I’ve long argued that these decisions belong in front of your Architecture Review Board, not just finance. In You Want to Migrate from VMware? Ask Your Architecture Review Board First, I made the point that the real barrier is architectural maturity, not budget line items. EVS makes that distinction even sharper.

The TCO Reality

• In VMC, the SRE function was baked in. In EVS, you inherit it.

• Your tooling and operational model remain “always-on,” with the same oversubscription patterns.

• Skill set variability matters — not every VMware admin team is prepared for AWS-style hardware churn.

If the architecture doesn’t change, the true TCO doesn’t change either. Broadcom has done this math. I’ve done this math. Moving to AWS under a new SKU doesn’t erase the operational and financial realities — it just changes who carries the pager.

My Take

EVS isn’t about cost savings or escaping VCF. It’s another on-ramp to VCF everywhere. Broadcom’s strategy is to make VCF the consistent VMware stack across your public and private cloud footprint — on-prem, AWS, Oracle, Google, Azure.

For customers in this new landscape — whether they’re coming from VxRail or VMC on AWS — the real decision is:
Do I inherit the lift of running VCF on-prem, or do I run it with a hyperscaler — and if so, which one?

For most mission-critical VMware shops, EVS is less about “freedom” and more about inheriting fragility without the operational safety nets they’ve come to depend on.

What This Means for the Market

AWS EVS isn’t the start of something new — it’s a defensive play in a market that VMware, under Broadcom, is actively reorganizing.

• If you’re an AWS partner, understand EVS customers are likely to be in “stabilize” mode, not “modernize” mode — yet.

• If you’re another hyperscaler, sharpen your VCF story, because customers will compare OCVS, GCVE, AVS against EVS.

• If you’re a VMware customer, know that VCF is the long game. Whether you run it yourself or let a vendor manage it is now your most critical architecture decision.

Keith on Call
If you’re staring down the same crossroads as former VMC on AWS or VxRail customers — deciding whether to take on the lift of running VCF yourself or pick a hyperscaler to run it — I can help.

With Keith on Call, you send me the details of your situation asynchronously, and I’ll send you a tailored response based on what I’m seeing in the market and in the field. No meetings, no scheduling — just a straight answer from someone who’s been deep in these decisions.

Because in this market, the wrong choice doesn’t just cost money — it costs uptime. Send me an email to inquire [email protected]