.png)
How It Works
A Resource-Based CUD is a billing commitment you make directly with Google Cloud. You agree to use a defined amount of compute resources, specified by machine type, vCPU count, memory, or GPU configuration, in a particular region for either one or three years. In exchange, Google applies a discounted rate to that usage automatically. Unlike the Spend-Based CUD, which commits to a dollar amount of spend, the Resource-Based CUD commits to a specific resource shape. That specificity is where its higher discount rate comes from. AWS calls the equivalent product Reserved Instances; Azure calls it Reservations.
Why It Matters for Cloud Cost
Resource-Based CUDs offer Google Cloud’s deepest per-resource discounts, making them valuable for teams running stable, predictable workloads on consistent machine types. The trade-off is rigidity. If your team changes instance families, migrates to a different region, or scales down significantly, the committed resources may go partially unused. Unused commitment still costs money. Without active monitoring and right-sizing before committing, teams routinely over-commit and pay for capacity they never consume.
Key Characteristics
- Commitments are scoped to a specific region and cannot be moved to another region after purchase.
- Discounts apply automatically at billing time with no action required from engineering teams.
- Resource-Based CUDs can be combined with Sustained Use Discounts on eligible workloads, but only up to the committed amount.
- Commitments cover vCPU and memory independently, so partial configurations are possible on some machine families.
How Usage AI Handles This
Usage AI’s CoPilot analyzes your GCP Compute Engine usage patterns and surfaces Resource-Based CUD recommendations with projected savings before any commitment is executed. Autopilot goes further, purchasing and adjusting commitments daily on your behalf, with cashback guaranteed on any underutilized commitment so you carry zero financial risk.
See how Usage AI saves 30 to 50% on AWS, GCP, and Azure.
Common Questions
1. How is a Resource-Based CUD different from a Spend-Based CUD?
A Resource-Based CUD commits to a specific machine type or vCPU and memory configuration, which is why it offers a deeper discount. A Spend-Based CUD commits to a dollar-per-hour spend across a broader range of compute services and trades some discount depth for flexibility across instance families.
2. What happens if I don’t fully use my Resource-Based CUD?
Google charges you for the committed resource amount regardless of actual usage. If your workload shrinks or changes shape mid-term, you still pay the committed rate on any unused portion. Usage AI’s Guaranteed Buyback protection covers underutilization through cashback credits, so you are not left absorbing that cost.
3. Can I combine a Resource-Based CUD with other GCP discounts?
Yes, within limits. Resource-Based CUDs stack with Sustained Use Discounts on eligible usage up to the committed resource amount. Beyond that threshold, Sustained Use Discounts apply to additional on-demand usage separately. You cannot apply both a Spend-Based CUD and a Resource-Based CUD to the same resource unit simultaneously.