返回
原文
The Decoder：AI News（RSS）
64AI 编辑部评分，满分 100

Nvidia 为自家芯片残值提供担保，以解锁 5000 亿美元 AI 基建融资

2026-08-11 17:41· 1小时前· Maximilian Schreiner
AI 导读

Nvidia 与 Apollo、BlackRock 等六家金融机构签署意向书，计划调动超 5000 亿美元第三方资本建设数据中心等设施。为促成融资，Nvidia 将为自家芯片残值提供最高 25% 的担保，承担部分折旧风险。CEO Jensen Huang 称 A100 商用六年仍在使用，并以 H100 租金上涨作为芯片长期价值的依据。

原文 · 未翻译
Image description

Key Points

  • Nvidia and six major financial firms plan to mobilize over $500 billion in third-party capital to build AI infrastructure like data centers.
  • To back the financing, Nvidia is guaranteeing up to 25 percent of the residual value of its own chips installed in these projects.
  • While critics warn of a financial bubble driven by rapidly obsolete hardware, Nvidia CEO Jensen Huang points to rising rental prices and argues that his processors have long economic lifespans.

Apollo, BlackRock, Blackstone, Brookfield, Goldman Sachs, and KKR are set to mobilize over $500 billion for AI infrastructure. To make the math work, Nvidia is guaranteeing a portion of the residual value of its own chips. Critics like investor Michael Burry consider the long-term value of those chips one of the biggest weak spots in the AI boom.

Nvidia has signed letters of intent with six major financial firms to mobilize more than $500 billion in third-party capital for data centers, chip factories, and power plants. The partners are Apollo, BlackRock, Blackstone, Brookfield, Goldman Sachs, and KKR. According to the Financial Times, which broke the deal, Nvidia's stock dropped about 1.4 percent afterward, wiping out more than $70 billion in market cap.

Nvidia CEO Jensen Huang described the move on X as a shift from one-off projects to repeatable financing platforms. "AI factories" should be fundable as productive infrastructure, similar to power grids or transportation networks. Many AI companies have demand for compute but can't access capital at the scale they need. Huang was clear that the $500 billion is an aggregate target spread over years, not Nvidia revenue, not a single fund, and not a commitment to any one customer. Nvidia didn't share terms, individual commitments, or a timeline.

Nvidia backs up to 25 percent of its chips' residual value

Huang also addressed accusations that Nvidia's financing of neoclouds and AI companies is circular. The company plans to support individual projects with residual-value guarantees. If the resale or reuse value of installed hardware falls below expectations at the end of a financing term, Nvidia covers part of the gap, up to 25 percent of a given transaction, reviewed on a project-by-project basis. The chipmaker is essentially taking on some of the depreciation risk on its own products. Huang says this share is "significantly lower" than in other compute financing arrangements, and that the actual credit assessment, meaning evaluation of the customer, demand, utilization, cash flow, and residual value, stays with the capital providers.

Nvidia regularly supports its partners in taking on debt, which in turn drives the company's own revenue. The company is also negotiating a guarantee for a 10-gigawatt data center in Ohio leased to OpenAI.

Huang's argument reads like a direct response to Michael Burry's depreciation warning

Huang's reasoning maps closely onto criticism from investor Michael Burry, who called the hyperscalers' depreciation practices "one of the more common frauds of the modern era." Burry argued that GPUs become obsolete too fast for five-to-seven-year useful lives because of Nvidia's two-to-three-year upgrade cycle, and that depreciation would be understated by roughly $176 billion between 2026 and 2028 alone.

Huang now claims the opposite. He says the A100, launched in 2020, is still in commercial use six years later and that its economic lifespan stretches toward a decade. CUDA, he argues, keeps improving installed hardware over time. As market evidence, he points to rising rental prices. H100 annual contracts went from $1.70 per GPU-hour in October 2025 to $2.35 in March 2026, while B200 capacity runs between $5.30 and $7.05.

Morgan Stanley expects hyperscaler spending of $3.5 trillion between 2026 and 2028. Apollo president Jim Zelter puts the total investment need at over $8 trillion. The Bank of England warned in its July Financial Stability Report that the pace is historically unprecedented and that a shock hitting highly leveraged AI companies could ripple through global financing conditions and trigger a credit crunch. Banks and private credit firms, the report noted, have limited visibility into their indirect exposure.

Nvidia

Financial Times / Nvidia deal

Jensen Huang / post on X

Bank of England / Financial Stability Report

行业动态

来源：The Decoder：AI News（RSS） · the-decoder.com

导出 Markdown
同一事件 · 2
返回
64打开原文
导出 Markdown

Nvidia 为自家芯片残值提供担保，以解锁 5000 亿美元 AI 基建融资

The Decoder：AI News（RSS）·2026-08-11 17:41·1小时前·Maximilian Schreiner
AI 导读

Nvidia 与 Apollo、BlackRock 等六家金融机构签署意向书，计划调动超 5000 亿美元第三方资本建设数据中心等设施。为促成融资，Nvidia 将为自家芯片残值提供最高 25% 的担保，承担部分折旧风险。CEO Jensen Huang 称 A100 商用六年仍在使用，并以 H100 租金上涨作为芯片长期价值的依据。

原文 · 保持原样，未翻译
Image description

Key Points

  • Nvidia and six major financial firms plan to mobilize over $500 billion in third-party capital to build AI infrastructure like data centers.
  • To back the financing, Nvidia is guaranteeing up to 25 percent of the residual value of its own chips installed in these projects.
  • While critics warn of a financial bubble driven by rapidly obsolete hardware, Nvidia CEO Jensen Huang points to rising rental prices and argues that his processors have long economic lifespans.

Apollo, BlackRock, Blackstone, Brookfield, Goldman Sachs, and KKR are set to mobilize over $500 billion for AI infrastructure. To make the math work, Nvidia is guaranteeing a portion of the residual value of its own chips. Critics like investor Michael Burry consider the long-term value of those chips one of the biggest weak spots in the AI boom.

Nvidia has signed letters of intent with six major financial firms to mobilize more than $500 billion in third-party capital for data centers, chip factories, and power plants. The partners are Apollo, BlackRock, Blackstone, Brookfield, Goldman Sachs, and KKR. According to the Financial Times, which broke the deal, Nvidia's stock dropped about 1.4 percent afterward, wiping out more than $70 billion in market cap.

Nvidia CEO Jensen Huang described the move on X as a shift from one-off projects to repeatable financing platforms. "AI factories" should be fundable as productive infrastructure, similar to power grids or transportation networks. Many AI companies have demand for compute but can't access capital at the scale they need. Huang was clear that the $500 billion is an aggregate target spread over years, not Nvidia revenue, not a single fund, and not a commitment to any one customer. Nvidia didn't share terms, individual commitments, or a timeline.

Nvidia backs up to 25 percent of its chips' residual value

Huang also addressed accusations that Nvidia's financing of neoclouds and AI companies is circular. The company plans to support individual projects with residual-value guarantees. If the resale or reuse value of installed hardware falls below expectations at the end of a financing term, Nvidia covers part of the gap, up to 25 percent of a given transaction, reviewed on a project-by-project basis. The chipmaker is essentially taking on some of the depreciation risk on its own products. Huang says this share is "significantly lower" than in other compute financing arrangements, and that the actual credit assessment, meaning evaluation of the customer, demand, utilization, cash flow, and residual value, stays with the capital providers.

Nvidia regularly supports its partners in taking on debt, which in turn drives the company's own revenue. The company is also negotiating a guarantee for a 10-gigawatt data center in Ohio leased to OpenAI.

Huang's argument reads like a direct response to Michael Burry's depreciation warning

Huang's reasoning maps closely onto criticism from investor Michael Burry, who called the hyperscalers' depreciation practices "one of the more common frauds of the modern era." Burry argued that GPUs become obsolete too fast for five-to-seven-year useful lives because of Nvidia's two-to-three-year upgrade cycle, and that depreciation would be understated by roughly $176 billion between 2026 and 2028 alone.

Huang now claims the opposite. He says the A100, launched in 2020, is still in commercial use six years later and that its economic lifespan stretches toward a decade. CUDA, he argues, keeps improving installed hardware over time. As market evidence, he points to rising rental prices. H100 annual contracts went from $1.70 per GPU-hour in October 2025 to $2.35 in March 2026, while B200 capacity runs between $5.30 and $7.05.

Morgan Stanley expects hyperscaler spending of $3.5 trillion between 2026 and 2028. Apollo president Jim Zelter puts the total investment need at over $8 trillion. The Bank of England warned in its July Financial Stability Report that the pace is historically unprecedented and that a shock hitting highly leveraged AI companies could ripple through global financing conditions and trigger a credit crunch. Banks and private credit firms, the report noted, have limited visibility into their indirect exposure.

Nvidia

Financial Times / Nvidia deal

Jensen Huang / post on X

Bank of England / Financial Stability Report

来源：The Decoder：AI News（RSS）· the-decoder.com

同一事件 · 2