Business•

Jensen Huang's Nvidia Exit Raises the Cost of Capital Question for Bitcoin Miners Building AI Data Centers

Jensen Huang's departure from Nvidia has put Bitcoin miners' AI buildouts under scrutiny, with JPMorgan's Erica Spear saying new deals could take longer or cost more.

Written By :MINRK
Jensen Huang's Nvidia Exit Raises the Cost of Capital Question for Bitcoin Miners Building AI Data Centers

Photo by Homa Appliances on Unsplash

Mentioned in This Article

BTC
BTC
Bitcoin
$64,686.00
0.19%

1. Nvidia's upheaval lands on the Bitcoin miners' AI pivot

Jensen Huang's resignation as Nvidia's chief executive and board member is testing the Bitcoin miners' AI pivot at the moment those companies have committed billions of dollars to it. Shares of IREN fell in the AI selloff of November 12, 2026, and JPMorgan credit analyst Erica Spear said new deals could take longer or cost more.

The timing is uncomfortable for a sector in transition. Over the past several years, companies that built their businesses on mining Bitcoin (BTC) have been redirecting power capacity, land and capital toward artificial intelligence infrastructure. Those plans depend on outside money and on customers with the means to pay for years of service. Nvidia sits close to the center of both, and uncertainty at the top of the chipmaker is now feeding into the question of how miners will pay for what they have promised to build.

No retreat in financing has been confirmed. What has changed is the degree of scrutiny. Investors are looking harder at which miners already have the money, the power and the contracts in hand, and which still need capital and regulatory approval before their expansion plans can become real.

2. Why does Nvidia's leadership change matter to Bitcoin miners?

It matters because Nvidia reaches the mining sector in three separate roles at once: as a supplier, as a customer and as a financier. Each of those relationships carries money into AI infrastructure projects, and a change in direction at the company could affect any of them.

The supplier role is the most familiar. Nvidia makes the graphics processors that AI data centers are built around, so any miner converting a site to AI work is planning around its hardware. The customer role is more recent and more direct. In May, IREN signed a five-year cloud services contract with Nvidia worth $3.4 billion, an agreement under which the chipmaker pays the former pure-play miner for computing capacity. A contract of that size and duration is the kind of commitment a company can show to lenders when it seeks funding for construction.

The third role is that of investor. Nvidia put $2 billion into CoreWeave, an AI cloud provider, in January. Cloud providers of that type are tenants for data center capacity, which makes their access to funding relevant to anyone hoping to lease them space. When a single company supplies the chips, buys the capacity and helps fund the firms that rent it, a shock to that company travels through the whole chain.

3. The questions investors are now asking

Two questions sit at the heart of the reassessment. The first is whether large AI contracts can produce enough cash flow to cover what it costs to build the facilities and to service the debt taken on to build them. The second is whether those contracts add up to a business that can sustain itself over time.

Neither has a simple answer. AI data centers require heavy spending before they generate revenue. A site has to be constructed, connected to power and filled with equipment, and much of that is paid for with borrowed money. A signed contract promises income, but the income arrives over years while the spending and the interest payments begin immediately. The value of a headline contract figure therefore depends on how it compares with the full cost of delivering the service.

Durability is the harder issue. A multi-year agreement provides visibility for its term, but a company that has reshaped itself around AI hosting needs the demand to continue after the first contracts run their course. Investors who once valued miners largely on how much Bitcoin they could produce are being asked to assess a different kind of company, with different economics and a different set of things that can go wrong.

4. Swapping Bitcoin price exposure for new kinds of risk

A miner's fortunes have traditionally risen and fallen with the price of Bitcoin. Revenue is earned in the asset, so a higher price lifts margins and a lower one squeezes them. Moving into AI infrastructure was attractive in part because contracted revenue from computing customers does not swing with a volatile market.

The pivot does not remove risk, though. It replaces one kind with three others: construction risk, financing risk and customer credit risk. Construction risk is the chance that a facility is delivered late or over budget. Financing risk is the chance that the money needed to finish a project becomes more expensive or harder to obtain. Customer credit risk is the chance that the party on the other side of a contract cannot meet its obligations for the full term.

Those risks do not fall evenly across the sector. A miner with funded projects, power it can actually use and payment commitments it can enforce is in a very different position from a competitor whose growth still depends on raising capital and securing regulatory approval. The first group has largely locked in the ingredients it needs. The second is exposed to every change in investor mood, and the turmoil at Nvidia is exactly that kind of change. The result may be a sharper split in performance between companies that have often been treated as a single trade.

5. How falling share prices make expansion more expensive

Equity markets delivered the first visible effect. IREN declined during the AI selloff on November 12 that followed Huang's departure, as investors marked down companies tied to the buildout of AI capacity.

A lower share price is more than a paper loss for a company that is still raising money. Miners have often funded growth by selling new stock, and the arithmetic of that approach turns against them when prices drop: to bring in the same amount of cash, a company has to issue more shares. Existing shareholders end up owning a smaller slice of the business, and each new round of fundraising becomes more costly in terms of ownership given up.

This is how a leadership change at a chipmaker can raise costs for a company that has not altered its own plans at all. The projects, the sites and the contracts are the same as they were before November 12. What has moved is the price at which the market is willing to supply the capital to complete them.

6. JPMorgan's Erica Spear sees slower or costlier deals

Credit markets are the other channel, and here the assessment from JPMorgan is cautious rather than alarmed. Spear, a credit analyst at the bank, said new deals could take longer to complete or could cost more. That is a statement about transactions still to come, not about agreements already signed.

The mechanism runs through Nvidia's own activity. If the chipmaker were to slow its investments, its guarantees or its purchases of capacity, miners could face higher financing costs and delays in construction. Lenders and investors take comfort from the presence of a large, well-capitalized counterparty. Less of that support would mean projects have to stand more fully on their own merits, and funding offered on that basis tends to be priced accordingly.

There is a second-order effect as well. Cloud providers that rent capacity from data center owners need funding of their own. If money becomes tighter for those tenants, their appetite for additional capacity could shrink, reducing demand for the very space miners are racing to build. A miner could therefore feel the pressure twice: once in the cost of its own borrowing, and again in the strength of the customers it hopes to sign.

7. AI data centers are not one business

Part of the difficulty in judging the sector is that the label covers several distinct activities. Leasing powered land, running colocation facilities and owning graphics processors in order to rent out computing are three different businesses, even when the same company describes all of them as an AI strategy.

They differ in what has to be built. A company leasing powered land provides a site and an electricity supply and leaves much of the rest to its tenant. A colocation operator builds and runs the facility in which a customer's equipment sits. A company that owns the processors and sells computing takes on the whole stack, including the most expensive hardware in the building.

They also differ in ongoing duties and in exposure to equipment replacement. Operating a facility brings continuing responsibilities that a landlord of powered land does not carry. Owning processors brings the risk that the hardware will need to be replaced as newer generations arrive, a cost that a company merely hosting someone else's machines does not bear. Each model therefore responds differently to a change in funding conditions, and a shock that barely touches one may weigh heavily on another.

8. A wider gap between funded and unfunded miners

Taken together, the events of the past several days sharpen a divide that already existed. On one side are miners whose AI projects are paid for, powered and backed by enforceable customer commitments. On the other are miners whose plans remain proposals until capital is raised and approvals are granted.

For the first group, uncertainty at Nvidia is mainly a question about future growth. For the second, it bears on whether current plans can be carried out on the terms originally expected. Higher financing costs, longer deal timelines and weaker share prices all fall hardest on companies that still need to go to the market.

The scale of the commitments explains why the distinction is drawing attention. Miners have pledged billions of dollars to AI infrastructure, and the returns on that spending depend on contracts performing as written and on funding remaining available at a workable price. Whether Nvidia continues to act as supplier, customer and financier in the way it has is now one of the variables investors are watching most closely.

Related Articles

NEWSLETTERS

Don't miss another story.

Subscribe to the MINRK Newsletter today.

Tick the newsletters you want. No selection? You'll get MINRK Headlines.

By signing up, you will receive emails about MINRK products and you agree to our terms of use and privacy policy.

MINRK Daybook

Market analysis for crypto traders and investors.

EVERY WEEKDAY

MINRK for Advisors

Defining crypto, digital assets and the future of finance for financial advisors.

EVERY THURSDAY

MINRK Protocol

Exploring the tech behind crypto one block at a time.

WEEKLY

MINRK Long & Short

A must read for institutions. Insights, news and analysis delivered weekly.

EVERY WEDNESDAY

MINRK Headlines

The biggest crypto news and ideas of the day.

EVERY WEEKDAY

MINRK Policy

Examining the intersection of cryptocurrency and government.

WEEKLY

MINRK Research

Join thousands of readers who rely on MINRK for data-driven insights on the latest digital asset trends.

MONTHLY