Nine public Bitcoin miners spent $5.11 billion on capital assets in the first half of 2026, while their AI and HPC business brought in just $341.2 million. The gap between spending and revenue reached roughly 15 to 1.
Across a broader group of 15 companies combining mining with AI infrastructure buildout, combined capital spending in their latest 2026 reporting periods hit $30.7 billion, up 42.6% from all of 2025. The figures come from consulting firm BlocksBridge in its weekly Miner Weekly newsletter.
Capital spending jumped 42% year over year
BlocksBridge calculated capital spending based on cash purchases and allocations to hardware, property, equipment and other productive assets, after subtracting proceeds from asset sales. The sample covers companies that combine classic Bitcoin mining with building out artificial intelligence and high-performance computing (HPC) capacity.
Throughout all of 2025, this group of 15 companies spent $21.53 billion. Already in the latest 2026 reporting period, capital spending has topped last year's full-year total by more than a third.
The capex acceleration coincides with a pullback in classic mining. According to related BlocksBridge data, public miners simultaneously cut combined hashrate by 13.4%, redirecting some of their power and capacity toward AI workloads.
A 15-to-1 gap between spending and revenue
Among miners that report AI and HPC revenue separately, the picture is even starker. Nine comparable companies spent $5.11 billion on capital assets in the first half of 2026, while combined AI and HPC revenue reached $341.2 million. That works out to roughly $15 in spending for every dollar earned from the new business line.
The gap is wide. But the trend inside it is already shifting.
AI revenue is accelerating, up 52% quarter over quarter
In the second quarter of 2026, the nine miners earned $205.8 million from AI and HPC, up 52% from the prior quarter. BlocksBridge named Core Scientific, TeraWulf and Bitdeer among the companies posting the biggest gains.
The progression is notable, though the total is still far from covering capital spending. These companies need years to reach breakeven on new capacity if revenue growth holds up.
Why the pivot to AI is so expensive
BlocksBridge explains why owning land and power contracts alone doesn't guarantee a fast move into the AI business. In the US, data center demand for electricity is growing faster than new grid capacity is being built. In that environment, access to already-interconnected sites and long-term power contracts has become a rare competitive edge for miners against newer entrants in the AI infrastructure race.
"Power contracts and available land may give miners a starting advantage, but converting those assets into AI-ready capacity requires substations, buildings, cooling systems, networking equipment and, in some business models, GPUs."
- BlocksBridge Consulting, Miner Weekly report, August 2026
That list explains why capital spending outpaces revenue by such a wide margin. Among the main cost items:
- Substations and higher power supply capacity
- Buildings and cooling systems for dense equipment layouts
- Networking infrastructure for data transfer
- GPUs: the priciest line item for models built on in-house compute
Bitcoin above $72,000 gives miners some breathing room
Bitcoin's price recovery is easing the pressure from another angle. The coin gained more than 13% over the past week and climbed back above $72,000 after the US Treasury said it would at least double the maximum size of its long-term bond buybacks, to $4 billion per operation. The move aimed to improve liquidity in the Treasury market and immediately boosted risk appetite.
The rally in BTC prices on global markets is also felt by Ukrainian users. Those looking to buy Bitcoin with hryvnia are paying noticeably more this week than they were seven days ago.
Whether the price recovery is enough to offset AI-related capital spending remains unclear. Most of these companies' core business is still classic mining, and its profitability depends directly on the coin's price.
Where the capital is heading
Markets have already reacted to this shift at the investment product level. CoinShares this week announced a strategy change for its industry-tracking ETF, rebranding it as the CoinShares Bitcoin Mining and Digital Power ETF (ticker WGMI). The fund, with $222.4 million in assets, now holds 29 positions spanning Bitcoin miners, data center operators, AI chipmakers, power generation companies and HPC providers.
A similar shift was already on display at CleanSpark, whose quarterly revenue missed Wall Street estimates despite its bet on AI contracts. Both cases show that the AI pivot gives miners a new growth avenue, but it hasn't replaced their core business yet.
The launch of WGMI and similar investment products shows that markets have already started valuing miners not just as Bitcoin producers but as infrastructure players in the AI industry. The coming quarters will show whether the 15-to-1 gap starts to close. For now, AI revenue is growing faster than capital spending in relative terms, even as the absolute gap remains large. For investors, this looks like a multi-year story rather than a quick turnaround.
Comments
Your email address will not be published. Required fields are marked *