Lambda raises $1 billion in debt to buy Nvidia chips it will lease to Microsoft

Lambda borrows $1 billion to buy Nvidia chips for Microsoft

Lambda, an AI cloud company that buys computing chips and rents the capacity to businesses, has raised $1 billion in private, short-dated debt to purchase Nvidia GPUs that it will then lease to Microsoft, according to Bloomberg. TechCrunch, reporting on the deal, says JP Morgan Chase arranged the financing.

The maturity is the telling detail. According to Bloomberg, the short-dated terms signal that Lambda expects to deploy the chips quickly and repay the loan just as quickly out of the revenue they generate. That makes the facility less like permanent capital and more like bridge financing backed by a committed customer — in this case, a major cloud operator.

The third GPU loan in four months

The Microsoft deal continues a pattern of borrowing against specific customer contracts rather than speculative capacity. In May, Lambda closed a $1 billion secured credit facility. Earlier this week, it announced a $926 million loan to fund Nvidia GB300 GPUs — one of Nvidia's newest chip models — for a deployment it is under contract to provide to Nvidia itself. Taken together, the three facilities amount to roughly $2.9 billion in new borrowing since spring, each tied to a named counterparty.

Equity fundraising in the background

The debt also lands amid Lambda's broader fundraising efforts. The company is reportedly in talks for a $3 billion pre-IPO round, and in November 2025 it raised $1.5 billion in venture capital at a $5.43 billion post-money valuation, according to PitchBook data cited by TechCrunch.

For a business that may be approaching the public markets, debt has a clear attraction: it expands the GPU fleet without diluting shareholders further, provided the contracts attached to each loan cover the repayments.

A $400 billion lending wave

Lambda is one participant in a much larger shift in how AI infrastructure gets paid for. According to data compiled by Bloomberg, banks and technology companies have raised more than $400 billion in AI-related debt globally so far in 2026. The buildout of data centers and GPU fleets has outgrown what equity funding and corporate cash can carry alone, and credit markets have become a main engine of the AI capital cycle.

Why it matters

The design of this loan says as much as its size. Lambda is not borrowing against a general growth story; it is borrowing against specific customers — Microsoft on this deal, Nvidia on another — who have committed to renting the very hardware the debt pays for. That lets so-called neoclouds scale capacity to signed demand instead of forecasting it, and it keeps dilution low ahead of a possible IPO.

The trade-off is concentrated risk. Short-dated debt secured on chips that depreciate quickly only works if deployments land on schedule and customers keep paying. And at the macro level, the $400 billion figure means lenders, not just venture investors, now share exposure to the AI cycle — if demand or GPU rental rates disappoint, the correction would run through credit markets as well as equity ones.