Bitcoin miners' move into AI is creating a new market for trading computing power
Luxor raised $16 million to expand into AI hardware and trading, betting that computing power will become a market where companies can hedge costs and prices.
Bitcoin mining companies are increasingly turning to artificial intelligence (AI), renting out computing power and converting mining facilities into data centers. Now, some of the companies that built their businesses serving bitcoin miners are following them into the AI market.
Luxor Technology, best known for its bitcoin mining services, is also moving into AI hardware and trading contracts that allow companies to agree on computing prices in advance.
The Seattle-based company said on Tuesday that it raised $16 million in a round led by CMT Digital. Bitcoin mining equipment makers Canaan and MicroBT also invested. Luxor said the money will go toward its AI hardware brokerage and computing-power trading business.
The round included new investment and sales of existing shares, although Luxor did not disclose how much of the new funding was invested in the company.
Publicly traded bitcoin miners had signed more than $70 billion worth of AI and high-performance computing contracts by the end of March, according to CoinShares data.
Not everyone is moving away from bitcoin mining. Nasdaq-listed NIP Group (NIPG) announced plans last week to essentially increase its mining capacity by about 48%. The company said it expects to generate about 195 bitcoin a month at full deployment.
Moving into AI also requires more than access to electricity. Ethan Vera, Luxor's chief operating officer, told CoinDesk that its software business works with bitcoin miners representing nearly two gigawatts of power capacity.
After assessing more than half that capacity, Luxor found that most sites had sufficient electricity and fiber connections for AI workloads but would require new data centers.
Luxor is also expanding into contracts that allow buyers and sellers to agree on future computing prices, much as bitcoin miners use financial contracts to manage changes in mining revenue.
Sam Hallene, a partner at CMT Digital, said AI companies have different reasons to protect themselves against price changes, including the possibility that expensive computing equipment remains useful longer than expected.
"We don’t know exactly what the market structure will look like at the end of the day," Hallene told CoinDesk.
Luxor already offers privately negotiated contracts tied to computing rental prices, although the market is still developing. Vera said transparency remains a problem, particularly around AI hardware prices.
In August, Luxor introduced a price index that tracks daily asking prices for Nvidia AI server systems, providing buyers and sellers with a reference point for equipment prices.
Diversified RWA stablecoins sustain 5-7% yield from real credit as crypto funding compresses to ~4%. GENIUS pushes yield off-chain; TAM grows to $4B in 3 years.
Asanat Analysis — Why it matters
Bitcoin miners are repurposing idle or underutilized hash rate infrastructure into tradeable commodities, a structural shift enabled by the fungibility of computing power across proof-of-work and AI workloads. Luxor's $16M raise signals institutional confidence in treating hashrate as a financial asset class with derivatives and spot markets—similar to energy futures or cloud computing credits. This matters because it creates a new revenue stream for miners during Bitcoin bear cycles and a cost-hedging tool for AI infrastructure operators, potentially compressing margins for both constituencies long-term.
Historically, mining has been a binary bet: run at full capacity or shut down. A tradeable compute market breaks that model, enabling miners to treat equipment as liquid financial positions rather than fixed industrial assets. The precedent matters: AWS and similar platforms already monetize spare capacity, but decentralized markets for hashrate could introduce price discovery mechanisms missing from centralized cloud providers. Risk: oversupply of tradeable computing power could commoditize AI infrastructure pricing, pressuring incumbents like Nvidia and major cloud providers operating at premium valuations.