Skip to content
Forkin

Live ledger

Our infrastructure emissions comparison

This page compares the Forkin core fleet with a disclosed reference cloud region. The cumulative CO₂e difference is modeled from fleet power assumptions and provider data; it is not a measured individual saving.

286

kg estimated CO₂e difference

Computed from conservative fleet electricity assumptions backed by provider and billing data, versus the counterfactual of running the same workload on a typical cloud region (AWS Frankfurt / us-east-1 grid carbon intensity).

This is a counterfactual difference between two estimates, not an individual saving or neutrality claim. Forkin is not carbon-neutral, buys no offsets, and reports residual estimated emissions separately. How this is calculated

814

kWh consumed

Conservative idle estimate — we round up our own draw, never down.

140

Days running

Since the first server in the fleet was provisioned.

7

Active servers

Each one logs its own hourly carbon row.

DataCrunch (Verda) · Hetzner · Hyperstack · Scaleway

Low-carbon providers

Hetzner, Scaleway, Hyperstack & Verda/DataCrunch today.

Where you fit

Your part in this

There is no per-user attribution math here - that would be dishonest and we'd rather not invent a number. What's true is simpler: every paid subscription helps pay for EU/EEA infrastructure selected for renewable contracts or low-carbon grids instead of cheaper grey-grid alternatives. The number above grows because you're here.

Methodology

How the number is calculated

Every hour, a cron job on our pipeline server appends one row per active server to a public ledger table. Each row is computed as:

kWh × (counterfactual_grid_intensity − our_provider_intensity)

Counterfactual grid intensities are 12-month averages from electricitymaps.com; our provider intensities come from provider-published data or the regional grid where a provider-specific figure is not available (for example Hetzner renewable contracts, Scaleway France's low-carbon grid, Hyperstack Norway hydropower, and Verda/DataCrunch Finland).

We use conservative fleet power assumptions, including idle draw where exact per-service metering is unavailable. The figure is the modeled difference from the counterfactual baseline above, not a measured saving or carbon-neutrality claim: we are not carbon-neutral, we buy no offsets, and we report the residual instead of cancelling it. The endpoint behind this page (/impact/totals) returns totals only.

Start

Scan a product on Forkin

The next scan you do uses these servers. The number above ticks up when you do.