Formula
8 GPUs × 10 hours = 80 GPU-hours
GPU-hours measure time-based access to accelerators, much like kilowatt-hours measure energy use over time.
Learn
What is a GPU-hour?
The basic unit behind compute pricing.
A GPU-hour is one GPU made available for one hour. It is a simple unit, but its market meaning depends on the chip type, memory, networking, region, commitment length, utilization, and service wrapper.
1 GPU × 1 hourUnit
The baseline way to express accelerator rental time.
Not fungibleCaution
One GPU-hour is not automatically equivalent to another GPU-hour.
Example
A simple GPU-hour calculation
If a workload uses 8 GPUs for 10 hours, it consumes:
Unit economics
What the unit measures
- Time-based access to an accelerator.
- A common way to discuss rental pricing across providers.
- A starting point for utilization, revenue, and capacity analysis.
Why it matters
Why GPU-hours matter
- They give buyers and providers a common language for compute usage.
- They are the starting point for comparing rental pricing across providers.
- They help translate hardware access into workload cost.
Market context
Why GPU-hours are not all the same
A GPU-hour tells you how long capacity is available, not how powerful that capacity is. Chip generation, memory, networking, region, commitment length, and bundled services can all change the value of one GPU-hour versus another.
- Chip generation and memory configuration matter.
- Networking, storage, support, and software stack can change effective value.
- Commitment length, region, and availability can move pricing even within the same chip family.
Keep learning
Related lessons
Concept
What is AI compute?
The basic resource behind training and running AI models.
Compare
H100 vs H200 vs B200
How accelerator generations affect performance, supply, and cost.
Market
What is compute cost?
How the market price of AI compute capacity is expressed and compared.