# Common questions

> Short answers to the questions readers ask most about the prices, the charts, the contracts and the site.

Updated 2026-09-06. Canonical: https://gpuquant.com/docs/faq

## Why is the H100 price here higher than the price I see advertised elsewhere?

Because the reference price is the hyperscaler list price, and the specialist GPU clouds sell the same hardware for roughly a third of it. GPUQuant publishes both: the reference price from the hyperscalers and the neocloud index from the specialist clouds, side by side and never averaged together. Large hyperscaler customers also negotiate below list, which no published figure can see.

[The neocloud index](https://gpuquant.com/docs/neocloud-index)

## Is this real-time?

The benchmarks are not, and should not be. The underlying prices are published price lists that change in discrete steps, sometimes staying still for a year. Ingestion runs daily, so a change is seen within a day, but the published resolution is monthly, because that is the finest the sources support. The Silicon Data indices are daily, and the instruments on Tradable Compute are priced live by their venues.

[Time and coverage](https://gpuquant.com/docs/history)

## Why is a chart flat for months at a time?

Because the price did not change. A list price that holds is the normal case, not a loading failure. Charts hold the last published value and mark held months in the tooltip.

[Held prices](https://gpuquant.com/docs/history#held)

## Why does a provider line simply stop, or never appear?

Because that provider publishes no eligible offering for that GPU. A missing provider is absent, never zero and never interpolated. The reason is shown in the legend and on the provider comparison table.

[Missing providers](https://gpuquant.com/docs/history#missing)

## Why is the twelve-month change sometimes N/A when the chart clearly shows a price?

Because the two endpoints were not built from the same providers. If a provider entered the series during the year, the difference between the endpoints would partly measure that change in coverage. GPUQuant rebuilds both endpoints from the providers present in both months, and when none span the period it declines to quote a number.

[Period changes](https://gpuquant.com/docs/history#period-changes)

## Why only H100 and B200?

They are the two GPUs CME Group and Silicon Data plan to list rental-index futures on. GPUQuant prices and stores several other NVIDIA datacenter GPUs every day, and publishing one is a matter of flipping a flag, but a benchmark that claims to be about tradable compute should track the tradable compute.

[Published and tracked](https://gpuquant.com/docs/reference-price#published)

## Is the GPUQuant Reference Price what the futures settle against?

No. The contracts settle against Silicon Data indices, SDH100RT for H100 and SDB200RT for B200, which GPUQuant displays under licence and does not compute. The GPUQuant figures are list-price benchmarks calculated independently from published prices.

[Silicon Data indices](https://gpuquant.com/docs/silicon-data)

## Does GPUQuant let me trade compute futures?

Not yet. The CME H100 and B200 futures list on October 5, 2026, pending regulatory review, and will trade through a connected futures broker that keeps the account. What trades today is on the Tradable Compute page: perpetuals on the venue named on each row, with orders signed in your own wallet and executed by the venue. GPUQuant is non-custodial, holds no funds and publishes no futures prices.

[Tradable Compute](https://gpuquant.com/docs/tradable-compute)

## Do I need an account?

Not for the data. Everything published is public, with no login and no per-user data. Trading uses a wallet, which is the account at the venue; GPUQuant keeps no account of its own.

[How a trade executes](https://gpuquant.com/docs/tradable-compute#execution)

## What are the figures at the top of the landing page?

Two published estimates of the scale of the build-out, each with its source printed under it, and a clock modelling the capital expenditure of four companies from their reported totals. None of them is a GPUQuant measurement.

[Landing page figures](https://gpuquant.com/docs/landing-figures)

## Can I cite or reuse a number from the site?

A GPUQuant figure, yes, with attribution and a link to the page it came from; every one is traceable to a published price and the method is on these pages. Silicon Data figures are shown under a licence that covers display on this site only, so they must be taken from Silicon Data directly.

[What the site shows](https://gpuquant.com/docs/silicon-data#licence)