Cipher Mining Expands AI Data Center Push With ERCOT Capacity And Long-Term Lease

• Cipher Mining disclosed conditional ERCOT interconnection capacity for up to 300 MW.
• The company also reported a 15-year hyperscaler master lease tied to HPC data-center infrastructure.
• The SEC filing does not identify the hyperscaler tenant by name.

Cipher Mining is pushing further into high-performance computing and AI infrastructure, using the power assets built around its mining business as a platform for longer-term data-center contracts.

A September 21 SEC filing outlines conditional ERCOT interconnection capacity for up to 300 MW alongside a 15-year hyperscaler master lease.

The combination matters because power access has become one of the most valuable assets in both Bitcoin mining and AI data centers.

Mining companies already control large power connections, land and electrical infrastructure. As demand from AI workloads increases, some operators are trying to convert part of that footprint into higher-value hosting arrangements.

Cipher’s filing places it firmly inside that trend.

The 300 MW figure refers to conditional ERCOT interconnection capacity, while the 15-year term applies to the disclosed hyperscaler master lease.

One detail should not be filled in with speculation.

The SEC filing does not name the hyperscaler tenant, so attaching a specific technology company to the agreement would go beyond the source.

That leaves the strategic direction as the more useful part of the story: Cipher is building an infrastructure business that can serve computing demand beyond Bitcoin mining.

For miners, long-term data-center leases can also change the revenue profile by reducing dependence on Bitcoin price and network difficulty.

The September 21 filing does not mean mining disappears from Cipher’s business. It shows the company is increasingly trying to monetize the same power and site-development capabilities across both crypto and AI/HPC markets.

This article was written by the News Desk and edited by Samuel Rae.

Source: Primary Source