4 hrs ago
Nvidia's Chip-Financing Plan Faces Wall Street Reality Check
Nvidia wants to help companies get money to buy or use powerful chips for artificial intelligence.
The plan could involve using those chips as collateral for loans.
Nvidia believes its best chips can keep earning money for up to 10 years.
Many banks and investors are less confident and usually assume the chips will lose value faster.
They want stronger guarantees or dependable customer contracts before lending large amounts.
This matters because many companies are spending huge sums on AI data centers and equipment.
Some early chip-backed loans have been supported by payments from companies such as Meta.
Nvidia says independent financing partners will judge each deal based on its expected cash flow and remaining value.
Nvidia’s proposed $500 billion financing initiative would use AI chips as collateral to help developers access computing capacity.
Banks and credit investors question whether advanced GPUs can retain revenue-generating value for the decade Nvidia projects.
Lenders are seeking stronger guarantees, customer contracts, or repayment protections than Nvidia initially outlined.
Nvidia says its AI computing assets are durable, productive, fungible, and capable of supporting long-term financing.
Existing GPU-backed deals have generally relied on dependable customer revenues or substantial guarantees from technology companies.
- Who
- NVIDIA Corporation, its financing partners, banks, credit investors, and AI companies seeking access to computing capacity.
- What
- A proposed $500 billion financing initiative is being reassessed because lenders doubt whether Nvidia’s AI chips can serve as long-term collateral.
- Where
- The debate is taking place among Wall Street lenders and investors financing AI infrastructure.
- When
- The initiative was announced in August, while financing deals are currently being structured and evaluated.
- Why
- Nvidia wants to create a financing market that helps AI developers obtain computing capacity without relying only on conventional subscription arrangements.
Nvidia’s view
Wall Street’s cautious view
Useful life of AI chips
Nvidia’s view
Nvidia says its GPUs and related AI computing assets can remain productive and generate revenue for up to a decade, citing third-party studies and valuations.
Wall Street’s cautious view
Banks and credit investors commonly underwrite GPUs over three to four years and lack enough historical data to confidently assign long-term residual value.
Strength of collateral
Nvidia’s view
Nvidia describes AI compute as a durable, productive, and fungible asset that can support long-term financing.
Wall Street’s cautious view
Lenders do not yet treat the chips as investment-grade collateral comparable to aircraft and want stronger guarantees or dependable customer revenues.
Structure of the financing plan
Nvidia’s view
Nvidia says the initiative brings independent institutional capital into AI infrastructure and is intended to address concerns about circular financing.
Wall Street’s cautious view
Investors are scrutinizing the use of vendor financing, private credit, and arrangements that could resemble circular financing, while demanding greater compensation and repayment protection.
Key facts
- Proposed financing size
- $500 billion
- Nvidia’s projected chip life
- Up to 10 years for some GPUs
- Typical bank depreciation schedule
- Three to four years
- Potential residual-value guarantee
- Nvidia initially said some deals could have guarantees of no more than 25%
- CoreWeave GPU-backed facility
- $8.5 billion, supported largely by Meta’s contractual payments
- Broadcom-backed financing
- Broadcom backstopped more than 80% of a $35 billion financing structure
- Nvidia financing partners mentioned
- Blackstone, Apollo, and KKR
Quotes
Tony Trzcinka
Senior portfolio manager at Impax Asset Management
“As investors, you’re going to be a lot pickier about the levels that you need to get compensated for to take incremental risk.”
CNBC TV 18
“Wall Street is much more conservative.”
CNBC TV 18
Nvidia spokesperson
Representative of Nvidia commenting on the financing plan
“AI compute is a productive, durable and fungible asset that can support long-term financing. Our financing partners independently assess each opportunity, including customer commitments, expected cash flow and residual value.”
CNBC TV 18
Sources
Related news
Broadcom to Lend Anthropic Up to $42 Billion for Chips
Anthropic Discusses $100 Billion IPO With Nvidia Investment
Nvidia’s AI Equity Stakes Surge to $99 Billion
Nvidia’s AI Investment Portfolio Poised to Surpass $100 Billion
Anthropic AI Warning Pressures Chip Stocks, But Trade Endures
AI Leaders Debate Slowdown as Investment Plans Keep Growing
India Needs Specialized Financing for GPU-Powered Data Centres
Silicon Valley Keeps Moving the Goalposts on AGI Claims
Dan Ives: AI revolution only 10-15% complete, biggest phase ahead
