Contents
Key takeawaysELA versus per productWhat we saw in 2024 and 2025Where the breakeven sitsRisks an ELA addsTerms that decide the numbersWhat Broadcom will sayChecking your core countWhat to do nextFAQA Broadcom ELA pays off only at scale. Below roughly 5,000 committed cores it rarely beats per product VMware licensing, and above that the discount holds only if a renewal cap and true down rights come with it.
- Scale decides it. ELA economics start working around 5,000 committed cores; below that, per product subscriptions keep your flexibility and your pricing easy to read.
- The discount depends on consumption. ELA rates hold only if you deploy what you committed, because idle cores raise the price of every core you actually use.
- Shelfware reverses the result. Unused ELA entitlements turned several deals in our benchmark from a saving into a loss.
- The renewal is the trap. A deep first term discount sets up an uncapped renewal on a VMware footprint that has become more dependent in the meantime.
- Per product keeps your exits open. Licensing cluster by cluster keeps a partial migration credible, which strengthens your hand at every renewal.
- Terms beat rates. Renewal caps, true down rights and divestiture clauses are worth more than extra discount points.
How do a Broadcom ELA and per product licensing differ?
A Broadcom ELA commits you to a multi year spend across a named set of products, in return for a deeper discount and one set of terms. Per product licensing buys VCF, VVF and attach SKUs cluster by cluster at street rates. The ELA gives up flexibility to get a better rate, and per product does the reverse.
Under Broadcom the distinction became sharper. The portfolio is already consolidated into a few per core bundles, so the simplification an ELA used to bring to the old VMware catalog of separately priced products is worth less today. What is left is a trade of rate against commitment.
What does each model commit you to?
- ELA. A fixed multi year spend across named products, consumed against committed quantities, with growth often priced in advance. You pay for the committed cores whether or not you deploy them.
- Per product. Annual or multi year subscriptions per bundle and per cluster, renewed and adjusted at each cycle. A compute cluster on external storage can sit on vSphere Foundation while a cluster that needs NSX networking takes VCF.
- Hybrid. A smaller ELA covering the stable part of your environment, with per product subscriptions at the edges. Most large buyers should price this structure, even if they end up choosing one of the other two.
All three count per physical core, with a minimum of 16 cores per CPU. Our guide to VMware core licensing explains the counting rule, and the VCF and VVF comparison covers which bundle each cluster needs.
Is Broadcom's Portfolio License Agreement the same as a VMware ELA?
No. Broadcom's Portfolio License Agreement covers its Enterprise Software catalog, mostly former CA products, in full or by segment such as DevOps or AIOps, and is sold on predictable annual costs. A VMware ELA commits quantities of VCF, VVF and add ons, so its value depends on deploying them.
The Broadcom ELA agreements guide and our CA mainframe licensing guide cover the portfolio side.
The VMware Estate After the Repackaging
What have we seen in Broadcom ELA evaluations in 2024 and 2025?
The breakeven was consistently higher than the sales model claimed. I ran roughly 20 to 30 Broadcom ELA versus per product evaluations for clients between 2024 and 2025, and three patterns kept coming back:
- Breakeven set too low. ELA proposals put breakeven at about half the core count our consumption analysis supported.
- Shelfware by mid term. Roughly 1 in 3 signed ELAs carried 15 to 30 percent shelfware by the middle of the term, which erased the headline discount.
- Terms decided the outcome. Buyers who negotiated true down rights and renewal caps kept the ELA economics. Buyers who negotiated only the rate lost them.
In roughly 8 of those evaluations, the per product or hybrid path cost less over five years despite a worse headline rate. Each of those ELA cases combined 15 to 30 percent shelfware with an uncapped renewal that took back every dollar the first term had saved.
Broadcom Enterprise Agreements Guide
ELA structures, discount benchmarks and exit clauses for Broadcom enterprise agreements, in one download.
Get the white paper →Where is the breakeven between a Broadcom ELA and per product?
In our 2024 to 2025 evaluations the realistic breakeven sat around 5,000 committed cores, with stable or growing consumption. Below that level the ELA discount did not survive shelfware risk and renewal exposure. Above it, the rate advantage compounded year after year.
Build the comparison on certified cores and conservative growth, never on the vendor forecast. The ELA proposal assumes the growth that justifies the commitment, so a model built on that forecast proves the case by assuming it.
| Factor | Broadcom ELA | Per product |
|---|---|---|
| Discount | 25 to 45 percent below street at scale | Standard negotiated rates |
| Flexibility | Locked to committed quantities | Adjustable at each renewal |
| Shelfware risk | High if growth disappoints | Low, scoped to what is deployed |
| Renewal exposure | One concentrated cliff at term end | Spread across cycles |
| Exit credibility | Weak during the term | Partial migration stays credible |
| Admin overhead | One agreement, one true up | More renewals to manage |
How should you model the comparison fairly?
Price three scenarios over five years: the ELA as proposed, per product at negotiated rates, and the hybrid. A three year window stops before the renewal, and the renewal is where the vendor recovers the first term discount, so write your renewal assumption into each scenario.
Then stress test the ELA against a 20 percent consumption shortfall. If it still wins with a fifth of the committed cores idle, the commitment is sized sensibly. If it wins only at full consumption, it has been sized to the forecast.
A worked example: 6,000 cores over five years
Say you run 6,000 cores today and Broadcom proposes an ELA for 7,500 cores over a three year term, sized on its forecast of 25 percent growth. Every figure here is hypothetical and chosen to show the mechanics:
- Per product rate. $200 per core per year, held flat for all five years.
- ELA rate. $130 per core per year, 35 percent below the per product rate.
- Actual growth. You stay at 6,000 cores for five years, so 1,500 committed cores sit idle, a fifth of the commitment.
- Hybrid. 5,000 stable cores committed at $150 (25 percent below the per product rate), plus 1,000 volatile cores bought per product at $200.
| Scenario | Years 1 to 3 | Years 4 and 5 | Five year total |
|---|---|---|---|
| Per product, 6,000 cores at $200 | $3,600,000 | $2,400,000 | $6,000,000 |
| ELA as proposed, renewal uncapped (7,500 cores renew at $200) | $2,925,000 | $3,000,000 | $5,925,000 |
| ELA with a 5 percent renewal cap and a true down to 6,000 cores at renewal ($136.50) | $2,925,000 | $1,638,000 | $4,563,000 |
| Hybrid, 5 percent cap on the committed block ($157.50) | $2,850,000 | $1,975,000 | $4,825,000 |
In the first term the idle cores lift the ELA's effective price from $130 to $162.50 per core in use, which is $975,000 a year spread over 6,000 cores. The 35 percent discount on paper is worth 18.75 percent in practice.
The renewal terms account for most of what remains. Without a true down right or a cap, assume Broadcom renews all 7,500 cores at the per product rate, which shrinks the ELA's five year saving to $75,000, about 1.25 percent. With both terms, the same ELA saves $1,437,000, roughly 24 percent, and the hybrid saves $1,175,000.
How does the answer change with the size of your environment?
Core count sets the starting point and your five year plans adjust it. The per core subscription calculator helps with a first pass.
| Your situation | Usually fits | What to watch |
|---|---|---|
| Under 5,000 cores, stable | Per product | Put compute only clusters on VVF and ask for price holds on renewal instead of a commitment |
| Under 5,000 cores, growing fast | Per product | Ask for a fixed rate on added cores without committing to them |
| Above 5,000 cores, stable or growing | ELA or hybrid | Sign only with the renewal cap and true down written into the order |
| Any size, planning a partial move off VMware | Per product or a small hybrid | Keep every cluster you may migrate outside the commitment |
| Group that buys or sells businesses | Hybrid, or an ELA with entity change terms | Carve out rights for divestitures and the ELA rate for acquired cores |
What risks does an ELA add that per product avoids?
An ELA adds three structural risks: shelfware from committed cores you never deploy, a renewal cliff where everything reprices at once, and a weaker exit position while the term runs. Broadcom's portfolio changes, visible in its public announcements, can also reshape what a bundle contains partway through the relationship.
The renewal cliff deserves the most attention. A three year ELA at a deep discount renews against a counterparty that has watched your VMware dependence grow for three years and knows precisely how locked in you are. Per product subscriptions can be staggered, so no single renewal date puts your whole VMware footprint on the table.
- Exit plans. Moving a cluster off VMware only cuts your bill if it is outside the commitment. Our note on using an exit plan in negotiation shows how to cost one.
- Renewal cliff. The Broadcom VMware renewal survival kit sets out how to prepare for the date the whole commitment reprices.
Why "take the ELA for the biggest discount" is the wrong starting point
Standard sourcing advice says large buyers should always take the ELA because its discount percentage cannot be beaten. We disagree. The discount is the one number Broadcom controls completely, while consumption and renewal assumptions are where you win or lose. In the 8 evaluations where per product or hybrid won, the ELA had the better rate every time.
Negotiate the terms that protect the downside first: the renewal cap, true downs and carve out rights. Once those are agreed, let the discount choose between structures that have both survived a fair five year model.
An ELA with a renewal cap and true down rights can be a good contract. An ELA with neither is a deferred price rise with a discount sticker on it.
Which ELA terms decide whether the numbers hold?
Four terms carry the economics: a renewal cap, true down rights, divestiture and entity change language, and clean audit clauses. They are worth more than extra discount points, and more than whatever the bundle adds to your entitlements on the Broadcom support portal.
- Renewal cap. A written maximum increase that applies at term end as well as during the term.
- True down rights. Scheduled checkpoints where committed quantities can fall to match what you deploy.
- Divestiture clauses. Carve out rights, so a sale or spin off does not strand entitlements you are still paying for.
- Audit terms. Verification against your own records, reasonable notice, and scoped tooling.
What contract wording should you ask for?
Put each of these in the order form or an amendment signed with it. An email from the account team will not bind the renewal team three years later.
| Term | Ask for | Why it matters |
|---|---|---|
| Renewal cap | Renewal price for the same products and quantities capped at a stated percentage over the final year of this term | Without it, the first term discount only postpones the increase |
| True down | An annual checkpoint where committed cores can drop to the certified count, with fees reduced from the next billing period | Turns shelfware from a sunk cost into one you can correct |
| Divestiture | Right to transfer or end entitlements for a divested entity, with a matching fee reduction | Otherwise you keep paying for cores the new owner runs |
| Acquisitions | Cores of an acquired company join at the ELA rate for the rest of the term | Stops an acquisition from opening a separate negotiation at street rates |
| Product changes | Equivalent rights at no added cost if Broadcom renames, rebundles or withdraws a committed product | Broadcom has reshaped the VMware bundles since the acquisition, and the ELA fixes your spend for the term |
Our guides to price cap negotiation and contract red lines go through these clauses one by one.
Is a hybrid structure negotiable in practice?
Yes, at scale. A committed block for the stable part of your environment, with per product subscriptions for volatile workloads, is a recognized structure. Pricing both paths in parallel is what makes Broadcom offer it, because the account team can see you have a costed alternative to the full commitment.
What will the Broadcom account team say, and how should you answer?
Expect the ELA to be presented as the natural choice for a buyer of your size. Typical lines, with replies:
- "The ELA rate is only available on a full commitment." Ask for the same rate on a smaller committed block, and table the per product quote you are comparing it with.
- "Your growth plans justify committing above today's count." Commit to the certified count and ask for added cores at the ELA rate as a price hold. Growth you never reach should not be paid for in advance.
- "This price expires at quarter end." Broadcom's fiscal year ends on the Sunday closest to October 31, so quarters close around early February, May, August and November. Use the date once both structures are modeled; see our note on quarter end timing.
- "We do not offer renewal caps." Then judge the proposal over five years with the renewal at street rates, and share that model with the account team.
- "True downs defeat the purpose of an ELA." Reply that consumption pays for the discount, and a checkpoint that matches the commitment to deployment is what keeps the agreement renewable.
How do you check your own core count before you model?
Start from your own inventory and reconcile the quote to it. Broadcom publishes a License Counting PowerCLI Tool in its knowledge base article 313548. Its FoundationCoreAndTiBUsage module reports licensable cores per host with the 16 core minimum applied, along with the vSAN capacity you use.
- PowerCLI counting tool. Run it against every vCenter. Broadcom asks for all physical cores to be active when the script runs, so check hosts with cores disabled in BIOS.
- RVTools. The vHost tab lists sockets and cores per host and gives you a second view to reconcile against the script.
- vSAN capacity. VCF includes 1 TiB of vSAN per core and VVF 0.25 TiB per core. Compare that with what you store before you accept add on capacity.
- Entitlements. Download current entitlements from the Broadcom support portal and match them to hosts, marking clusters due to retire within five years.
Which mistakes turn an ELA discount into a loss?
- Committing to the forecast. Signing for the cores the growth plan predicts instead of the cores you run. In the worked example, idle cores cut the discount on paper almost in half.
- Counting small hosts at face value. A two socket host with 12 core CPUs has 24 physical cores and is billed for 32. Consolidate or retire small hosts before you set the committed quantity.
- Committing clusters you plan to leave. Once they sit inside the ELA, migrating them saves nothing until the term ends.
What to do next
- Certify the core count. Run the counting tool, reconcile it with RVTools, and separate stable workloads from volatile ones.
- Build the five year model. Price the ELA, per product and hybrid, with the renewal assumption written into each.
- Stress test the ELA. Rerun it with a 20 percent consumption shortfall and an uncapped renewal.
- Set the term sheet. Agree the renewal cap, true down rights, divestiture language and audit terms before you discuss price.
- Price both structures in parallel. Ask Broadcom to quote the ELA and per product side by side, so it competes against itself.
- Time the signature. Aim for a Broadcom fiscal quarter end, with your model finished well before the date arrives.
Frequently asked questions
At what size does a Broadcom ELA make sense?
When you run well over 5,000 cores and expect steady or rising use. Under that line, idle committed cores and the renewal exposure usually cost more than the discount saves. If you sit near the line, price a hybrid with a committed block for your stable clusters before choosing either extreme.
What discount does a Broadcom ELA deliver?
Typically 25 to 45 percent below street rates at real scale, and only while you consume what you committed. Measure it as the price per core you actually deploy: with a fifth of the committed cores idle, a 35 percent headline discount is worth under 20 percent in use.
What is the biggest risk in signing a Broadcom ELA?
The renewal cliff. Every committed core reprices on the same day, against a vendor that knows how dependent you have become, which makes an uncapped renewal the costliest clause in most ELAs we review. Write the cap percentage into the order form itself.
Can you negotiate true down rights in a Broadcom ELA?
Yes, at scale and with parallel per product pricing on the table. Scheduled true down checkpoints were the single most valuable term in our benchmark files. Ask for an annual checkpoint tied to the certified core count, with fees reduced from the next billing period.
Is a hybrid ELA and per product structure realistic?
Yes. A committed block for stable workloads with per product subscriptions for volatile ones is negotiable, and it frequently cost the least over five years in our evaluations. It works best when the per product part holds the clusters you may retire or migrate.
Does per product licensing weaken your negotiating position?
It costs rate but keeps flexibility and a credible exit, and both strengthen your position at each renewal. One in three ELAs in our files lost its rate advantage to shelfware, a cost per product buyers avoid by paying only for deployed cores.