The email usually arrives a few months before renewal. A Saudi enterprise that has run VMware quietly for a decade — a couple of clusters, a stable estate, licenses bought once and supported annually — opens the new quote and finds a number that no longer resembles the old line item. Nothing about the workloads changed. The licensing world around them did.
This guide explains what actually changed after Broadcom's acquisition of VMware, why it lands differently in the Kingdom, and the four realistic paths forward. Spoiler: for most estates the answer is not a dramatic rip-and-replace — it is a deliberate sorting of workloads across paths.
What changed, in plain terms
All of the following is drawn from public reporting and analyses of Broadcom's licensing changes; exact terms vary by contract and date, so verify your own quote against current terms:
| Change | What it means |
|---|---|
| End of perpetual licenses (announced December 2023) | You can no longer buy VMware once and pay only for support; everything moves to subscription |
| Massive SKU consolidation | Thousands of individual products folded into a few bundles centred on VMware Cloud Foundation (VCF) and vSphere Foundation — you may pay for components you never deploy |
| Core-based minimums | Public reporting describes a minimum of 72 cores per order from April 2025, up from 16 — small estates pay for capacity they do not have |
| Renewal price impact | Analyses across the industry report multi-fold increases for many customers, with small environments hit hardest by the minimums |
| Late-renewal surcharges | Reporting during 2025 described surcharges applied to renewals past their anniversary date |
Individually, each change is a commercial decision a vendor is entitled to make. Collectively, they turned a stable cost line into a strategic question — especially for the smaller half of the market, which is precisely where most Saudi private-sector estates sit.
Why this lands differently in Saudi Arabia
Three local amplifiers make the Broadcom era more acute in the Kingdom than the global headlines suggest:
- Currency and budgeting. Subscriptions are USD-denominated; Saudi budgets are SAR-denominated and approved annually. A multi-fold USD increase mid-cycle is not just expensive — it breaks the budget process itself.
- Procurement rhythm. Enterprise and semi-government buyers work through tender and approval cycles measured in months. A licensing change that demands a decision before renewal compresses a twelve-month process into a quarter.
- The skills market. VMware administrators are plentiful; engineers who can run an alternative hypervisor estate in production are scarcer and in demand across the Kingdom's data-centre build-out. Any path that assumes instant re-skilling is fiction.
Path 1 — Stay and renegotiate
Right when: your estate is large, deeply integrated (SRM, NSX, vSAN), regulated, or mid-refresh — and the true cost of any migration exceeds the licensing delta.
The realistic tactics: consolidate onto fewer, denser hosts so the core count you license reflects reality; ruthlessly match the bundle to what you actually deploy; negotiate multi-year terms only at a discount that justifies the lock-in; and price the alternative paths anyway — a credible exit plan is your only real leverage in the room. Treat the renewal as a procurement event with alternatives, not an invoice to be paid.
Path 2 — Re-platform the hypervisor
Right when: virtualisation is core to your operating model, you have (or can hire) strong infrastructure engineering, and your hardware has years of life left.
The candidates Saudi teams shortlist are Proxmox VE (KVM-based, open-source, no per-core licensing), Nutanix AHV, Microsoft Hyper-V, and OpenStack/KVM or XCP-ng for particular profiles. The honest accounting: the hypervisor license is only one line. Add migration tooling and testing, backup-ecosystem compatibility, monitoring integration, staff retraining, and a period of running two platforms in parallel. Estates report the effort is very manageable for straightforward VM fleets and substantial for environments leaning on VMware's advanced networking and storage. Budget the project honestly before comparing it to the renewal quote — and if the numbers still favour re-platforming, they often favour it decisively, because the saving repeats every year.
Path 3 — Rehost to managed cloud
Right when: what you actually run on VMware is general-purpose — web applications, line-of-business apps, databases of ordinary size, file services, dev/test. For these, the hypervisor was never the point; it was the delivery mechanism. Moving the workload to a Saudi-based cloud server removes the licensing question entirely: no hypervisor to license, no hosts to refresh, SAR billing instead of USD subscriptions, and Saudi data residency into the bargain.
This path has a property the others lack: it can be tested this week. Stand up a VPS in a Saudi region, move one representative workload, and measure — performance, operations, cost per month. Skyline Cloud's 14-day free trial (no credit card) exists for exactly this kind of evidence-gathering; a fortnight of real data beats a quarter of committee slides.
What does not belong on this path: workloads pinned to owned specialised hardware, or systems whose regulator or client contractually requires owned equipment — those belong in the colocation column of our colocation vs cloud framework.
Path 4 — The hybrid shrink (what most estates actually do)
The most common real-world outcome is not one path but a sorting exercise: keep a small VMware core for the handful of systems that genuinely need it (and negotiate hard on that reduced footprint), rehost the general-purpose majority to managed cloud, and re-platform selectively where engineering appetite exists. The renewal quote shrinks with the core count; the risk spreads across paths instead of concentrating in one program.
| Estate profile | Sensible default |
|---|---|
| < ~20 VMs, general-purpose | Path 3 — rehost to managed cloud; the minimums make small VMware estates the hardest hit |
| Mixed estate, some advanced VMware features | Path 4 — shrink the core, move the rest |
| Large, VCF-integrated, regulated | Path 1 now, Path 4 over the refresh cycle |
| Strong infra team, owned young hardware | Path 2 — re-platform and keep the metal |
Migration guardrails, whatever the path
Sequence by risk, not by size: move stateless and easily reversible workloads first. Take verified backups before every move and test restores — a migration is precisely when backup and disaster-recovery discipline earns its keep. Keep rollback capacity available until each workload has survived a full business cycle (month-end, campaign peak). And update your cost model as you go: the point of the exercise, from our cloud cost guide, is a bill you can predict again. For teams that need hands, our parent company's server infrastructure practice works both sides — sustaining what stays, moving what goes.
Frequently asked questions
Do I have to leave VMware?
No — and many large estates should not, at least not this cycle. The mistake is neither staying nor leaving; it is renewing by default, without pricing the alternatives. Leverage requires a credible plan B.
Are the price increases really as large as reported?
Impacts vary enormously by estate size, bundle fit and negotiation; public analyses describe increases from meaningful to multi-fold, with small environments most exposed to core minimums. Ignore the headlines and price your renewal against your alternatives.
Is an alternative hypervisor risky for production?
The mainstream alternatives run serious production workloads worldwide. The real risk is organisational — skills, backup ecosystems, operational maturity — which is why Path 2 suits teams with engineering depth and Path 3 suits everyone else.
Can Saudi data residency requirements survive a move to cloud?
Yes — that is one of the quiet wins. Rehosting to a Saudi-region provider keeps workloads in-Kingdom; see our PDPL hosting guide for the contract questions that make it stick.
Turn the renewal shock into a sorting exercise
List your VMs, mark which ones genuinely need VMware, and price the other three paths for everything else. Half of that exercise costs nothing to run: open a free Skyline Cloud trial (14 days, no credit card), rehost one real workload to a Saudi cloud server, and walk into your renewal negotiation with evidence instead of anxiety.
Comments
0 total · 0 threads