Build, Install, Buy or Rent? The Economics of Getting an AI Factory
The price of a GPU-hour is the wrong number to compare. What matters is the cost of each hour a GPU does useful work, and that depends on how busy you keep it, how long it stays useful and what you pay for power, space and people. This final part puts current prices next to the costs of owning, lets you build your own break-even model, and turns the series into a buyer's scorecard.
1. The right question
A GPU you own costs roughly the same every hour whether it is busy or idle: the hardware is depreciating, the building and staff are paid for. A GPU you rent on demand costs nothing when you are not using it. So the comparison is not "owned price per hour" against "rental price per hour", but cost per useful GPU-hour: total cost divided by the hours that actually do work. Utilisation decides almost everything, which is why it came up in every part of this series and in CoreWeave's loan arithmetic in Case Study 7.
2. What renting costs
The price ladder
Indicative per-GPU-hour prices from trackers in 2026. Toggle the chip. Ranges are wide and change monthly.
The same H100 can rent for anywhere from about $1.49 to $6.98 an hour, a 4.7-times spread, with market medians around $2.95โ3.46. AWS's eight-GPU H100 instance works out to about $6.88 per GPU on demand but nearer $4 with commitments. B200 prices are more volatile: trackers in late 2026 range from about $5 to $8.40, and one index reported a 79% rise in three months as demand for the newest chips outran supply.
3. What owning costs
Where an owned GPU's money goes
Annual cost per GPU under the calculator's default assumptions (section 4). Hardware dominates.
- Hardware, spread over its useful life: the biggest line, and the most uncertain because the useful life of a GPU is debated (Case Study 7).
- Power: about 1.8 kW per GPU including cooling overhead, from one estimate of 1,000 H100s drawing about 1.76 MW, at US commercial rates around $0.12โ0.14 per kWh. See the power bottleneck.
- Space: US retail colocation of about $200โ400 per kW per month in primary markets, wholesale $80โ157 for blocks above 1 MW; or the cost of your own building.
- People: the specialists described in Who Runs the AI Factory.
- Software and support: platform licences, support contracts and spares.
4. The calculator
Cost per useful GPU-hour, four ways
Our simplified model, with every assumption editable. "Buy & run" = own hardware in colocation with your own team. "Install" = a cloud runs its hardware in your building (modelled as owning cost plus a management premium, minus most of your staff). "Reserved" = rent capacity for the full term. "On-demand" = pay only for hours used.
5. The break-even curve
Cost per useful hour vs utilisation
Uses the calculator's inputs. Owning gets cheaper per useful hour the busier you are; on-demand stays flat. Where the lines cross is your break-even utilisation.
6. Beyond cost
Most of the customers in Part 5 did not decide on cost alone. A ministry that must keep data in the country will pay more to do so. A hospital needs answers next to its scanners. A startup without a GPU team should not hire one to save money on paper. A telco may accept a long payback for a strategic business. And a cheaper route that locks you into one provider for five years may cost more later.
7. Hybrid is normal
IDC's survey of more than 1,000 decision-makers (Part 1) found that most organisations expect to need both on-premises and public cloud for AI. In practice the split tends to follow the workload.
Which workload goes where
A common pattern, not a rule. Tap a workload.
8. Buyer's scorecard
The four routes, re-weighted your way
Scores 1โ5 are our editorial summary of Parts 1โ5. Set how much each criterion matters to you.
9. The series
10. Reading the numbers
- Rental prices
- H100: $1.49โ6.98 range, medians ~$2.95โ3.46; AWS ~$6.88 on demand, ~$4 committed (CloudZero, Spheron, DataStorage). B200: ~$5โ8.41 across trackers; +79% in three months per one index (Shattered citing SemiAnalysis); Mercatus average $6.39. Reserved discounts "up to 60%" are a vendor-blog claim.
- Owning inputs
- Colocation $200โ400/kW-month retail primary, $80โ157 wholesale (Encor and others); electricity ~$0.12โ0.14/kWh; ~1.76 MW per 1,000 H100s (Spheron). Hardware cost, useful life, staffing ratio, software cost and the install premium are our assumptions.
- Model
- Straight-line depreciation, no financing cost, tax, resale value or price changes over time. On-demand assumes capacity is always available when needed, which is not guaranteed for the newest chips.
- Scorecard
- Editorial.
- Prices
- CloudZero: H100 cost in 2026
- CloudZero: cloud GPU pricing comparison
- Spheron: GPU cloud pricing 2026
- Ornn: GPU cloud price index
- Mercatus: GPU rental prices by provider
- DataStorage: GPU rental prices, mid-2026
- Shattered: B200 price surge
- Owning
- Encor Advisors: colocation pricing
- Datacentres.com: wholesale vs retail colocation
- Spheron: AI inference electricity cost
- Introl: colocation for 120 kW racks
- Part 1: the four routes