The first miner has already run towards AI

[safar1980]

Bitcoin’s Great Unplug: $1.5 Billion in Hardware Behind the AI Pivot

The hashrate public bitcoin miners shed in the first half of 2026 carries an uncomfortable price tag: roughly $1.5 billion in equivalent hardware investment, even at an assumed acquisition price of just $20 per terahash (TH/s).

A previous Miner Weekly issue estimated that public miners lost 75 EH/s of realized hashrate while their directly reported HPC and AI revenue rose 52% quarter over quarter. Much of that retreat reflected a redirection of power toward AI infrastructure.

The revenue growth showed what operators were gaining. But putting a purchase price on the disappearing hashrate helps illuminate what they had already paid for.

At $20 per TH/s, 75 EH/s corresponds to $1.5 billion of mining machines. That excludes the buildings, electrical equipment, cooling and installation needed to operate them.

That is an often-overlooked cost of the AI transition: investors had already financed the mining capacity being displaced. At some sites, equipment was marked down within months of starting production, raising questions about how much of that investment mining could recover before AI took over.

[philipma1957]

Bitcoin’s Great Unplug: $1.5 Billion in Hardware Behind the AI Pivot

The hashrate public bitcoin miners shed in the first half of 2026 carries an uncomfortable price tag: roughly $1.5 billion in equivalent hardware investment, even at an assumed acquisition price of just $20 per terahash (TH/s).

A previous Miner Weekly issue estimated that public miners lost 75 EH/s of realized hashrate while their directly reported HPC and AI revenue rose 52% quarter over quarter. Much of that retreat reflected a redirection of power toward AI infrastructure.

The revenue growth showed what operators were gaining. But putting a purchase price on the disappearing hashrate helps illuminate what they had already paid for.

At $20 per TH/s, 75 EH/s corresponds to $1.5 billion of mining machines. That excludes the buildings, electrical equipment, cooling and installation needed to operate them.

That is an often-overlooked cost of the AI transition: investors had already financed the mining capacity being displaced. At some sites, equipment was marked down within months of starting production, raising questions about how much of that investment mining could recover before AI took over.

Yeah years ago when it was  purchased it may have been 1.5 billion.

But if it is all s19 gear current value is closer to 2 dollars a th or 150 million . 

And the logistics  are expensive to sell it

[ucrem]

I agree that the main common asset between Bitcoin mining and AI/HPC is power availability.

But I think it is important not to confuse access to hundreds of MW with having a data center that can simply be converted from mining to AI.

A Bitcoin mining facility is normally designed around one objective:

Code:

minimize total cost per TH/s

This means that anything which does not directly contribute to running the ASICs tends to be reduced as much as possible.

You need power distribution, transformers, basic networking, cooling/ventilation and enough infrastructure to keep the miners running, but there is usually little economic reason to build the level of redundancy and facility infrastructure required by a modern enterprise or HPC data center.

AI/HPC is a completely different engineering problem.

Depending on the availability target, an AI data center may require:

- redundant utility feeds;
- UPS systems;
- generator capacity;
- A/B power distribution;
- redundant switchgear;
- fault-tolerant cooling;
- high-density or liquid cooling;
- fire detection and suppression;
- multiple fiber carriers;
- high-capacity internal networking;
- physical security;
- continuous monitoring;
- maintenance without service interruption.

And modern AI racks can also have power densities that are completely different from traditional data center designs.

So when a mining company says it is moving toward AI, I would not interpret that as:

Code:

mining facility -> replace ASICs with GPUs

It is closer to:

Code:

existing site with large power allocation
        +
grid interconnection
        +
land
        +
possibly reusable electrical infrastructure
        |
        v
build a largely new data center around those assets

The power contract and grid connection may actually be the most valuable part of the mining site.

That is also why miners can have an advantage entering this market: obtaining hundreds of MW of new grid capacity can be much harder and slower than buying servers.

But the existence of that power capacity does not remove the enormous CAPEX required to build a proper AI/HPC facility.

One small terminology point: I would also separate "AI data center" from "Tier IV data center".

Tier IV is a specific availability/fault-tolerance classification. A very large AI/HPC facility does not necessarily need to be Tier IV, but even without Tier IV certification its infrastructure requirements are still dramatically higher than those of a typical Bitcoin mining site.

So in my opinion Bitcoin mining and AI are competing primarily for energy and grid capacity, not because the facilities themselves are interchangeable.