Nvidia explores insurance-backed financing for AI chip customers

Traders Union
All News
Financial News
Nvidia insurance-backed chip financing

Highlights

Nvidia is in early talks with insurers and capital providers, including Howden Re, to shift lending risk on AI chips for neocloud customers.
Nvidia recently backstopped $500 billion of Wall Street financing and guaranteed $105 billion in leases for OpenAI's major data centre project.
Forward Compute cites Barkr AI research showing Nvidia's eight-GPU H100 system from 2022 retains near $320,000 value, dropping to about $30,000 after six years.

Financing structures for neocloud demand

As first reported by Financial Times, Nvidia has held talks with insurance companies about arrangements that would shift part of the risk of lending against its chips to insurers and other capital providers.

One proposal under discussion involves insuring losses on loans to smaller cloud computing groups, often called neoclouds, if they default and the Nvidia chips pledged as collateral cannot be resold for enough to repay lenders. The conversations remain at an early stage and may not result in deals, but they show the chipmaker is testing new ways to broaden the pool of buyers able to finance its hardware.

Nvidia chief executive Jensen Huang has argued that chips should be treated as an investable asset class, similar to other expensive long-life equipment that supports structured financing markets. Nvidia recently offered to backstop part of financing arrangements aimed at unlocking $500 billion of capital from Wall Street firms including Goldman Sachs and Apollo, and it also guaranteed $105 billion of leases tied to a large data centre project for OpenAI.

People familiar with the discussions said Nvidia has shared data on chip depreciation and expected future computing prices with at least one insurer. Another person said the company is working with broker Howden Re on a structure involving insurers, while also exploring ways to syndicate risk to hedge funds and other alternative investors because the possible deal size could exceed even large insurers' balance-sheet capacity.

Insurance market support for AI infrastructure

The discussions come as insurers roll out products aimed at the AI infrastructure build-out, including cover for credit losses, declines in chip values and contract breaches linked to power outages or cooling failures at data centres. Nvidia said its AI infrastructure is an investable asset class because it is productive, durable and fungible, and that capital partners help expand compute access for AI-native companies.

The company has also examined structures similar to residual value insurance, which protects against declines in the value of technology equipment. Providers such as Forward Compute and American Compute are offering products in this area, with Forward Compute saying insurance can reduce counterparty concerns that otherwise make large buyers reluctant to contract with smaller neocloud providers.

Valuation data is emerging as a key part of that market. Forward Compute said it is reviewing forward-looking chip valuations from research firms including Barkr AI and Silicon Data, while a forthcoming Barkr AI study estimates that Nvidia's eight-GPU H100 system from 2022 is still worth about $320,000 today, close to its initial value. The study projects that if computing supply catches up with demand, the system retains about two-thirds of its value after one year and falls to roughly $30,000 after six years.

Our earlier article on surging funding for AI inference startups highlighted how investor appetite is accelerating as companies that run trained models raise ever-larger rounds at sharply higher valuations. We noted Modal Labs’ talks to raise a major new round alongside similar moves by peers, underscoring strong demand for inference capacity even as margins remain pressured and operational risks like security incidents stay in focus.

This material may contain third-party opinions, none of the data and information on this webpage constitutes investment advice according to our Disclaimer. While we adhere to strict Editorial Integrity, this post may contain references to products from our partners.