IBM Db2 Licensing Metrics: Five Models, One Complex Choice
IBM Db2 is licensed through five distinct metrics, and the right choice depends heavily on your deployment environment, workload profile, and whether you're planning to use IBM Cloud Pak for Data. Selecting the wrong metric does not just result in overspending — it can also create audit exposure if your actual usage doesn't align with the metric's counting rules. Understanding IBM ILMT compliance is a prerequisite for any PVU-based Db2 deployment.
The five Db2 licensing metrics are: Processor Value Unit (PVU), Virtual Processor Core (VPC), Authorised User, Resource Unit (RU), and Cloud Pak entitlement (which bundles Db2 with other IBM Data and AI products). Each metric targets a different deployment scenario, and IBM's field teams have a habit of positioning the metric that maximises their revenue rather than the one that minimises your cost. Our IBM Cloud Pak licensing negotiation guide covers how to evaluate these choices with full pricing transparency.
PVU vs VPC: The Core Db2 Pricing Decision
PVU (Processor Value Unit) is IBM's legacy metric for Db2. On x86 hardware, each physical core is assigned 70 PVU; on IBM POWER systems, 120 PVU per core. Under sub-capacity PVU licensing, only the virtual cores allocated to Db2 VMs need to be licensed — not the entire physical host. A Db2 instance on a 2-vCPU VM on x86 hardware costs 140 PVU under sub-capacity, versus potentially thousands of PVUs at full physical capacity. This is why PVU sub-capacity compliance is so commercially important, and why IBM's ILMT requirement is non-negotiable for this metric.
VPC (Virtual Processor Core) is IBM's newer, Cloud Pak-aligned metric. One VPC equals one virtual core, regardless of the underlying processor family. The simplicity is genuine — no PVU tables, no ILMT requirement for VPC-licensed products, no sub-capacity/full-capacity distinction. However, IBM uses VPC primarily through Cloud Pak for Data, which bundles Db2 with Watson Studio, Watson Machine Learning, and other IBM data products. Enterprises that only need Db2 may pay for Cloud Pak components they'll never use. IBM has also eliminated volume discount structures from some PVU renewal paths, making standalone Db2 PVU renewals 10–20% more expensive than comparable VPC Cloud Pak deals in some cases. The economics require case-by-case analysis rather than a blanket recommendation.
Need a Db2 Licensing Cost Comparison?
Redress Compliance models PVU sub-capacity versus VPC Cloud Pak economics for your specific Db2 estate. Our clients consistently identify 20–40% savings opportunities that IBM's account teams never volunteered. 100% independent advisory — no commercial relationship with IBM. Talk to an IBM Specialist →
Db2 Authorised User, Resource Unit, and Cloud Pak Licensing
Authorised User (AU) licensing allows a specific named individual to access Db2 regardless of how many cores or virtual machines it runs on. AU is cost-effective only for small user populations — typically under 25 named users — and becomes less competitive as user counts grow. IBM requires separate AU licences for each person who accesses Db2 directly or through middleware, which makes AU licensing complex in large enterprises where Db2 sits behind application layers with many indirect users. The indirect access rules for Db2 Authorised User are less aggressive than SAP's equivalent rules, but they still require careful analysis.
Resource Unit (RU) licensing applies to specific Db2 workloads — primarily batch and analytics — and measures consumption based on workload throughput metrics rather than installed infrastructure. RU is relatively uncommon and typically applies to specific IBM product configurations rather than general-purpose Db2 deployments. Cloud Pak for Data entitlement is the most strategically significant Db2 licensing path in 2025–2026: it bundles Db2 Standard or Advanced under VPC, alongside IBM Watson Studio, Watson Machine Learning, and DataStage, using a shared VPC pool. Enterprises planning to use IBM watsonx alongside Db2 should model whether Cloud Pak for Data is more cost-effective than separate product licences, given that the VPC pool is shared across all included products.
Db2 on AWS, Azure, and Google Cloud
IBM Db2 is available on AWS Marketplace, Azure, and Google Cloud through BYOL (Bring Your Own Licence) and hourly metered pricing. BYOL requires valid IBM Db2 licences — typically under the Cloud Pak VPC metric, though some legacy PVU licences include cloud deployment rights under IBM's Passport Advantage licence terms. IBM's cloud deployment rights are not automatically included in all Db2 licence types: older PVU perpetual licences sometimes exclude cloud deployment unless Subscription and Support (S&S) is current, and even then the eligible cloud platforms are specified in the licence metric document.
AWS RDS Db2 — launched by AWS as a managed Db2 service in 2023 — uses a different pricing model: the Db2 licence cost is included in the RDS hourly rate, making it a consumption model rather than a capital licence. For enterprises looking to reduce complexity and eliminate ILMT management overhead, RDS Db2 is worth evaluating — though the per-hour cost at scale frequently exceeds BYOL under sub-capacity PVU. Book a call with our IBM team to model the full 3-year cost comparison for your specific Db2 workload, including ILMT management costs which are frequently omitted from IBM's own cost projections. For background on ILMT requirements, review our IBM ILMT Guide and download our IBM ILMT Audit Readiness resources.
Db2 Upgrade Paths and Migration Considerations
IBM Db2 version strategy has a direct impact on licensing costs. Db2 11.5 is the current release as of 2026, with extended support for Db2 10.5 and 11.1 available under IBM's software support lifecycle policy. Enterprises running Db2 10.1 or earlier should be aware that extended support costs — typically 40–60% surcharges on annual S&S — accumulate rapidly and often exceed the cost of migrating to current versions. IBM's account teams frequently use end-of-support pressure to push customers into Cloud Pak transitions, which may or may not be the most cost-effective path. A frank analysis from an independent adviser — not IBM's commercial team — is essential before making this decision.