A note to John Gowings · Gowings Bros · Compute, from the customer's side
Demand
What the customer is buying when they buy compute. The middle of the data flow is the only thing billed. What goes in and what comes back is the customer's own.
IWhat gets billed
Input
a prompt
a query
a document
a question
raw data
Process · Compute
GPU-hours of work done
the only thing billed to the customer
Output
a token
an answer
an image
a prediction
a model
USD 2–4per million output tokens at frontier;USD 1.50–3per H100 GPU-hour at retail.
If the customer owned the fleet capacity sized to peak
~30%average utilisation. The other 70 per cent is paid for, and idle.
Pooled across many peaks and troughs cancel
~70%operator utilisation. Your peak is everyone else's trough.
IIIWhen ownership pays, when renting does
Below ~67% utilisation
The chip is idle too often to amortise. The all-in cost of build divided by hours actually used exceeds the rental rate. Renting wins. Almost every customer lives here.
Above ~67% utilisation
You can amortise the chip; build undercuts rent. Constant workload (Meta serving global ad inference) or a strategic mandate (sovereign defence) is required. A handful of customers live here.
The investor read
Your peak is everyone else's trough. That arbitrage is what the data centre operator sells. The chips age in three years; the pooled demand had better not. Whether it does is the investor's question — the only real one once the connection is secured.