Loading prices…
STKR NewsSTKR News0 of 3 free this month
AI

Core Scientific revenue doubles in Q2 as AI colocation expansion accelerates

Core Scientific just posted a staggering $1 billion paper loss, but the underlying pivot from BTC mining to AI infrastructure tells a story of survival and massive growth.

Originally on Cointelegraph
AB

Adrian Boysel

Contributor

Jul 28, 2026

4 min read

Photo illustration / STKR News

The Pivot is Real

Core Scientific is currently walking a tightrope that defines the modern era of industrial-scale computing. The company recently released its second-quarter earnings, and the numbers are a masterclass in why you should never just read the headline net loss figures. On paper, they lost over $1.1 billion. In reality, they are transforming from a pure-play Bitcoin miner into an AI infrastructure powerhouse, and the markets are starting to pay attention.

The massive net loss mostly boils down to non-cash accounting charges related to their emergence from bankruptcy and the way they mark-to-market certain warrants. For builders and founders, this is the first lesson: GAAP earnings often mask the actual operational health of a business. When you strip away the accounting gymnastics, Core Scientific doubled its revenue compared to the same period last year. That doesn't happen by accident in a post-halving environment.

Mining vs. Hosting

For years, the narrative for companies like Core Scientific was simple: plug in as many ASICs as possible, hope the price of Bitcoin goes up, and try to keep your electricity costs low. But the halving changed the math for everyone. Margins for mining are thinner than they have ever been. Core Scientific saw the writing on the wall earlier than most, largely because their bankruptcy forced a radical rethink of their business model.

The company is now aggressively pivoting toward AI colocation. They aren't just mining blocks anymore; they are leasing out their massive data centers to house GPUs for high-performance computing. This shift is significant because AI contracts are typically much stickier and more predictable than the volatile rewards of Bitcoin mining. If you're building in the infrastructure space, this is a clear signal that the world is hungry for power and cooling, and it doesn't really care whether those electrons are securing a decentralized ledger or training a large language model.

The CoreWeave Factor

A huge part of this revenue jump comes from their massive deal with CoreWeave. By repurposing their existing infrastructure to support AI workloads, Core Scientific has secured a revenue stream that isn't dependent on the fluctuating price of BTC. This is a massive hedge. During the quarter, the company ramped up their capacity significantly, proving that the specialized power setups originally built for mining can, with enough capital expenditure, be converted for AI use.

However, this transition isn't cheap. Converting a site from a crypto mine to a Tier 3 data center requires a complete overhaul of the power distribution, cooling, and security systems. The builders at Core Scientific are essentially rebuilding the plane while it's in the air. The fact that their revenue doubled during this transition phase suggests that the demand for AI compute is currently outstripping the available supply of high-voltage power sites.

Power as the New Currency

In the world of crypto and AI, the real bottleneck isn't code or even chips—it's power. Core Scientific’s primary asset isn't their fleet of Bitmain miners; it's their access to the grid. In many parts of the United States, getting a new 500-megawatt connection can take five to ten years. Core already has the plugs. By pivoting to AI, they are essentially arbitrage-ing their power contracts. They are moving from a low-margin commodity (Bitcoin) to a high-demand service (AI hosting).

This should be a wake-up call for founders in the decentralized compute space. While we talk a lot about distributed GPU networks, the institutional money is flowing toward centralized, massive-scale facilities that can guarantee uptime for enterprise clients. Core Scientific is betting that having the physical real estate and the power permits is a moat that software cannot easily cross.

The Skeptic's View

It’s not all sunshine and high-speed chips. The $1.15 billion net loss, even if largely non-cash, highlights how messy these corporate restructurings can be. The company is carrying a significant amount of debt and the capital requirements to finish their AI build-outs are astronomical. If the AI bubble cools off before these facilities are fully operational, Core Scientific could find themselves overextended once again.

Furthermore, the operational complexity of managing AI data centers is much higher than mining. Bitcoin miners are "interruptible" loads—you can turn them off when power prices spike. High-performance computing for AI usually requires 99.9% uptime. You can't just flip the switch when the grid gets tight without breaching your service level agreements. This is a fundamental shift in how the company must operate at a technical level.

What Builders Should Take Away

If you are building in the crypto-AI overlap, there are three key takeaways from Core Scientific's quarter:

  • Infrastructure is King: Stop worrying about the token price and start looking at the physical constraints. The companies winning right now are the ones who controlled the power and the land years ago.
  • Diversification is Survival: Relying on a single revenue stream, especially one as volatile as mining, is a recipe for disaster. Core Scientific survived because they were able to pivot their existing assets to a new market.
  • Accounting Matters: For founders looking to scale, understand that your "net loss" doesn't always reflect your "free cash flow." Learning to communicate your operational wins despite accounting hurdles is essential for keeping investor confidence.

The Bottom Line

Core Scientific is no longer a Bitcoin mining company; it’s a power management and infrastructure company that happens to mine Bitcoin. The doubling of their revenue is a validation of the AI pivot, but the road ahead is paved with expensive construction and high expectations. They have demonstrated that they can secure the contracts, but now they have to prove they can operate at the level that AI giants demand. For the rest of us, it’s a reminder that in a gold rush, you want to be the one owning the land where the miners are digging.


Read the original at Cointelegraph →

The Brief

Stay Updated on Cutting-Edge Tech

A six-minute morning dispatch on the markets and the technology shaping them.

Free. No spam. Unsubscribe anytime.

Write for STKR

Become a Contributor

Earn $STKR for published stories on markets, protocols, and culture.

  • Earn $STKR for every published piece
  • Editorial support from the STKR desk
  • Byline visibility across the network
  • First look at the upcoming creator program
Apply to Write

Keep reading

All stories

Comments

24 reader responses