⏱️ vCPU Allocations to Compute Hours
vCPU → core-hour math with reserved instance commitment implications. 16 vCPUs × 730 hours × 10 instances = 116,800 vCPU-hours/month. AWS RIs and Savings Plans are sold in vCPU-hour commitments. Mismatch your estimate by 20% and you're buying on-demand at full rate for the gap. Model your fleet before signing.
📐 Fleet Profile
Set vCPUs/instance, instance count, and runtime hours. The engine computes vCPU-hours, vCPU-days, vCPU-months. Compare against common RI commitment tiers to spot the gap.
vCPU-Hours: The Universal Cloud Compute Unit
Every major cloud provider — AWS, Google Cloud, and Microsoft Azure — prices compute capacity using vCPU-hours as the fundamental billing unit. A vCPU (virtual CPU) represents a single hardware hyperthread exposed to a virtual machine instance. When you run an instance with 16 vCPUs for one hour, you consume 16 vCPU-hours of compute. This metric normalizes billing across instance families, regions, and operating systems, making it the lingua franca of cloud cost estimation.
How This Calculator Works
The calculator uses a straightforward multiplicative model grounded in how cloud providers meter compute usage:
| Formula | Definition | Example (16 vCPUs × 10 instances × 24 hr) |
|---|---|---|
vCPU-Hours = vCPUs × Instances × Hours | Total virtual CPU runtime across the fleet | 16 × 10 × 24 = 3,840 vCPU-Hours |
vCPU-Days = vCPU-Hours ÷ 24 | Equivalent full-day compute allocation | 3,840 ÷ 24 = 160 vCPU-Days |
vCPU-Months = vCPU-Hours ÷ 720 | Equivalent 30-day month (720-hour) compute allocation | 3,840 ÷ 720 = 5.33 vCPU-Months |
Fleet Sizing Across Cloud Providers
Each cloud provider exposes vCPU counts differently across its instance families. AWS EC2 instances range from 1 vCPU (t4g.nano) to 448 vCPUs (u-24tb1.112xlarge). GCP maps vCPUs to its committed use discount tiers, where sustained-use discounts automatically apply after a portion of the month. Azure expresses vCPU count in its vmSize property and meters per-minute granularity. By computing total vCPU-hours, you can normalize cost comparisons across providers, regions, and reservation models — a critical step when negotiating enterprise agreements or planning multi-cloud migrations.
Why vCPU-Hours Matter for Capacity Planning
vCPU-hours are not just a billing abstraction — they are a capacity planning tool. When you know your workload requires 50,000 vCPU-hours per month, you can reverse-engineer instance combinations: one 64-vCPU instance running 24/7 provides 46,080 vCPU-hours/month (64×24×30), leaving a gap of ~4,000 vCPU-hours that might be filled by spot instances or autoscaling bursts. Without this conversion, teams tend to over-provision by rounding up to the next instance size, wasting 20–40% of their compute budget.