CA mainframe under Broadcom, the savings are in the families you retire
Broadcom acquired CA Technologies in 2018 and inherited the largest non IBM mainframe portfolio, 50 plus products across security, storage, systems management, and development, priced on MIPS or the newer MSU sub capacity model. Mainframe customers are structurally captive because replacement means multi year refactoring, and Broadcom has applied 30 to 80 percent increases at renewal across documented enterprise cases, yet across our engagements the bill was driven by bundled families no one still used.
Prepared by Redress Compliance · August 8, 2026 · Broadcom advisory. Based on 20 to 30 Broadcom CA mainframe renewals advised 2024 to 2025.
Executive summary
The idle families are the bill: one or two carried real use while 30 to 50 percent of the bundle sat idle.
The portfolio spans four families, security with Top Secret and ACF2, storage and output with CA Disk, View, and Deliver, systems management with OPS/MVS, Workload Automation, and Sysview, and development with Endevor, Datacom, IDMS.
And Easytrieve, with most enterprises running 5 to 15 CA products priced separately. The largest savings come from retiring unused families, not from negotiating the headline MIPS rate, and the workload analysis is what separates the commercially critical products.
No alternative without years of refactoring, from the commercially negotiable ones.
The MIPS reconciliation is the true down nobody runs.
Broadcom meters CA subscriptions against contracted MIPS at signature with True Forward escalation when consumption exceeds it.
And the trap runs the other way: customers whose mainframe utilization dropped after distributed migrations kept the historical MIPS commitment and overpaid 15 to 30 percent, because mid term true down is not supported and the renewal is the only opportunity.
Reconcile actual deployed MIPS against contracted MIPS at every renewal, without exception.
The MSU transition is the 25 to 50 percent lever on the products that support it.
Broadcom is migrating the portfolio from MIPS to the MSU metric aligned with IBM's framework.
Measuring the rolling four hour average with sub capacity licensing of individual LPARs and workload groups rather than full machine capacity: customers on full capacity pricing with multiple LPARs and varying workloads captured 25 to 50 percent reductions through the sub capacity transition.
The reporting is the condition, running and archived, because gaps exposed one in three estates to full capacity pricing under audit.
The renewal preparation runs 12 to 18 months out and captures 20 to 35 percent.
Contracts run three to five year terms with annual billing, uplifts of 30 to 80 percent landed where capacity grew without a negotiated cap, and the disciplined sequence, deployment audit, MIPS or MSU reconciliation, workload analysis, alternative evaluation against IBM and BMC competing products.
And target outcome modeling, delivered 20 to 35 percent against Broadcom's opening proposals, compounding across the term.
The four renewal dimensions to work: scope, which products actually renew; quantity, the committed capacity; term length; and the bespoke structures at scale.
The two metrics, and the population you sit in
| Metric position | What it means | The buyer move |
|---|---|---|
| Under MIPS | Paying for capacity not used, common after distributed migrations | True down at renewal, the only window |
| At MIPS | Paying the contracted capacity | Hold, and cap the escalation |
| Over MIPS | Contracted plus True Forward escalation at premium | Reconcile before Broadcom does |
| MSU sub capacity | R4HA measured, LPARs licensed individually | The 25 to 50 percent transition where supported |
| Bespoke | Negotiated structures at upper scale | Price against the standard alternatives |
The MSU model aligns with IBM's meter, and the discipline pays twice.
The rolling four hour average, sub capacity LPAR licensing, and workload license charge categories mirror the IBM framework, so the same peak management that cuts the IBM MLC bill cuts the CA bill on migrated products, and the reporting obligation mirrors too: running and archived.
Or the audit defaults to full capacity, the exposure one in three estates carried.
The workload framework, critical against negotiable
- Map the dependencies: heavy CICS workloads lean on CA Workload Automation and Sysview; heavy DB2 workloads on Datacom and IDMS; the workload context decides the leverage.
- Classify commercially critical: products with genuinely no alternative short of years of refactoring, where the negotiation is about caps and terms, not exit.
- Classify commercially negotiable: products where IBM or BMC alternatives exist with manageable migration cost, the credible walkaways that move the whole proposal.
- Retire the idle: the 30 to 50 percent of the bundle carrying no use, whose removal is worth more than any rate negotiation on what remains.
The Broadcom enterprise agreements brief
The portfolio mechanics, the metric transitions, and the negotiation sequence across the Broadcom estate.
Get the white paper →The renewal sequence, 12 to 18 months out
The preparation runs in order: the deployment audit establishing which of the 5 to 15 CA products actually run and at what consumption; the MIPS or MSU reconciliation against contracted capacity, where the post migration estates found their 15 to 30 percent overpay.
The workload analysis separating critical from negotiable; the alternative evaluation against IBM mainframe products and BMC, priced seriously enough to be credible; and the target outcome model that turns it all into a position.
The captive structure is exactly why the preparation matters, Broadcom knows replacement takes years, so the leverage comes from scope reduction and the negotiable subset rather than exit threats on the critical one.
The newer MSPS portfolio packaging and its bundle mechanics are worked in the CA mainframe pricing analysis, the IBM side of the same estate in the mainframe CIO advisory, and the metric conversion pattern in the PVU to VPC transition guide.
- Percentile standing for your exact deal size and industry, from real closed transactions
- Scenario simulation before the call: test alternative terms and see the financial impact of each
- A negotiation playbook, talking points, and a two page executive brief on day one
What we saw across CA renewals, 2024 to 2025
Across roughly 20 to 30 Broadcom CA mainframe renewals Fredrik Filipsson advised between 2024 and 2025, the bill was driven by bundled families no one still used:
Of the bundle carrying no real use while one or two families did the work.
Estates open to full capacity pricing under audit through sub capacity reporting gaps.
The capacity growth reset is the pattern that costs most: renewal uplifts of 30 to 80 percent landed where the mainframe grew without a negotiated cap, because uncapped growth reprices the whole bundle including the idle families.
So the estate pays the increase on products nobody has opened in years.
The three disciplines compound in the buyer's favor instead: the deployment audit that strips the idle 30 to 50 percent from scope, the reconciliation that trues the capacity down to actual, and the cap that holds the trajectory.
And together they are the 20 to 35 percent the prepared renewals captured against the opening proposal.
Your first five moves
- Start the renewal sequence 12 to 18 months out: deployment audit, reconciliation, workload analysis, alternatives, target model.
- Retire the idle families from scope, the 30 to 50 percent of the bundle worth more than any rate negotiation.
- Reconcile deployed against contracted capacity, the true down window that only opens at renewal.
- Run the MSU sub capacity transition where products support it, the 25 to 50 percent on full capacity estates.
- Cap the uplift and archive the reporting, against the 30 to 80 percent resets and the audit default. The Broadcom practice runs the renewal with you.
Frequently asked questions
How does Broadcom license CA mainframe software?
On capacity metrics: MIPS, the legacy measure metering contracted capacity at signature with True Forward escalation above it, or the newer MSU model aligned with IBM's framework, measuring the rolling four hour average with sub capacity licensing of individual LPARs.
The portfolio spans four families, security, storage, systems management, and development, with most enterprises running 5 to 15 products priced separately.
How much do Broadcom CA renewals increase?
30 to 80 percent uplifts have been applied across documented enterprise cases since the acquisition, landing hardest where mainframe capacity grew without a negotiated cap, because uncapped growth reprices the entire bundle including idle products.
The disciplined 12 to 18 month preparation sequence delivered 20 to 35 percent against opening proposals in our engagements.
Can you reduce CA mainframe MIPS commitments?
Only at renewal: mid term true down is not supported, so customers whose utilization dropped after distributed migrations kept historical MIPS commitments and overpaid 15 to 30 percent until the next renewal window.
The reconciliation of actual deployed capacity against contracted belongs in every renewal preparation, without exception.
What is the MSU sub capacity opportunity on CA products?
25 to 50 percent on most products that support it: customers on full capacity MSU pricing with multiple LPARs and varying workload patterns pay for capacity they never use.
And the sub capacity transition licenses individual LPARs and workload groups at the measured rolling four hour average instead.
The condition is reporting, running and archived, because gaps exposed one in three estates to full capacity under audit.
Which CA mainframe products can be replaced?
The workload analysis decides: products deep in the critical path, source control in Endevor, security in Top Secret or ACF2 on heavy estates, are commercially critical with no alternative short of years of refactoring, while others have IBM or BMC alternatives with manageable migration cost.
The negotiable subset priced credibly is the walkaway that moves the whole proposal, and the idle 30 to 50 percent needs no replacement at all, just retirement.
When should CA mainframe renewal preparation start?
12 to 18 months before expiry: the deployment audit, the MIPS or MSU reconciliation, the workload analysis, the alternative evaluation, and the target outcome model take that long to run properly, and the 20 to 35 percent they captured compounds across a three to five year term.
Starting at the quote means accepting the quote's frame, scope, capacity, and baseline included.
