Nvidia turns to insurers to spread the risk of AI build-out

Nvidia has held talks with insurance companies about shouldering the risks of lending against its chips as CEO Jensen Huang pushes to unlock more demand beyond Big Tech. The chipmaker has approached insurers about structures shifting capital-intensive semiconductor financing risk, including coverage for losses on loans to “neoclouds” if they default and pledged chips cannot be resold. The discussions are at an early stage but show Nvidia experimenting across Wall Street, private capital, and insurance to widen its customer base.

Huang has described chips as an investable asset class like aeroplanes. Nvidia offered to backstop financing to unlock $500bn from Goldman Sachs and Apollo, guaranteed $105bn of leases for an OpenAI data centre, and expects a quarter of next year’s revenue from AI labs it supports on its balance sheet.

Insurers are launching AI infrastructure products covering credit risk, chip value falls, and contract breaches from power or cooling failures. Nvidia shared depreciation data with at least one insurer and is working with broker Howden Re, led by Ingemar Lanevi, head of financial solutions. The scale may overwhelm insurers, so Nvidia is exploring consortiums with hedge funds and asset managers.

Nvidia已与保险公司举行谈判，就将承担借贷其芯片风险一事进行探讨，因为CEO黄仁勋正推动在大型科技公司之外 unlock 更多需求。芯片制造商已向保险公司探讨将资本密集型半导体融资风险转移的方案，包括在“neoclouds”违约且无法转售已质押芯片时承担损失的保险。尽管讨论处于早期阶段，但这显示Nvidia正在尝试在华尔街、私募资本和保险业之间拓宽客户群。

黄仁勋将芯片比作飞机等可投资资产类别。Nvidia提出为背书融资以解锁来自高盛和阿波罗的5000亿美元，并为OpenAI数据中心承担1050亿美元的租赁担保，预计明年将有四分之一的收入来自其平台上支持的AI实验室。

保险公司正推出覆盖信用风险、芯片价值下跌以及因供电或散热故障导致合同违约的AI基础设施产品。Nvidia至少与一家保险公司共享了折旧数据，并正与Howden Re合作，该公司由Ingemar Lanevi领导，担任财务解决方案负责人。规模可能会让保险公司不堪重负，因此Nvidia正与对冲基金和资产管理公司组建联盟。

关键实体：公司：Nvidia、OpenAI、高盛、阿波罗、Howden Re | 人：黄仁勋、Ingemar Lanevi、Lee Harris、Ryan Mc Morrow、Roula Khalaf | 地点：伦敦、旧金山、台北、台湾

常见问题：为什么Nvidia让保险公司介入芯片融资？

为了将贷款和租赁的风险转移给保险公司，鼠

The user is asking me to translate a English source text into multiple languages. The user has also provided me with specific instructions:

1. Translate into all requested languages: zh-Hans, es, hi, ar, fr, bn, pt, ru, ja, de, id

2. For each language, output ALL fields: title, seo_title, seo_description, full content, faq_q, I need to translate the English source text into multiple languages. The user has provided me with specific instructions:

1. Translate into all requested languages: zh-Hans, es, hi, ar,1, a: experiences岛 (way, the, way, the text, all这些一个