Full story
Nvidia’s $500B financing push
Nvidia has partnered with Apollo, BlackRock, Blackstone, Brookfield, Goldman Sachs, and KKR to commit up to $500 billion for AI data center construction, with the larger aim of creating a secondary market for aging GPUs.
“Nvidia has agreed to guarantee, with its own money, that its chips used as collateral in these deals will retain their value”
The plan is structured around Nvidia agreeing to guarantee that GPUs used as collateral will retain their value, covering up to 25% of the difference if the chips fail to hold their expected resale value.
Data Center Knowledge says Zayo is building more than 8,000 miles of long-haul fiber with anchor customer Nvidia and will build six new long-haul routes while expanding capacity across 10 high-demand markets.
Zayo’s program spans more than 15,000 route miles across North America, and its CEO Steve Smith said in a statement that "AI is fundamentally reshaping where and how network infrastructure needs to be built across the U.S."
Analysts quoted by Data Center Knowledge add that pump lasers used in amplifiers and coherent transponders face near-term constraints, and that "I think both are constrained" for fiber and optical equipment.
Collateral risk and China
CNBC risk analysis cited by Tech Times frames the biggest threat to Nvidia’s collateral as China’s rapidly expanding domestic AI chip manufacturing base, warning that it could flood the global secondary GPU market with lower-cost silicon.
Ben Emons, founder of FedWatch Advisors, is quoted saying, "Depreciation is the one key risk here," and he adds that Nvidia chips could depreciate faster than expected.
Tech Times also describes Nvidia’s financing mechanics as requiring borrowers to deploy Nvidia-specified hardware architectures, standardizing collateral and giving lenders a recovery path if a borrower defaults.
Plataforma Media reports Huang’s framing that Nvidia’s AI factory platform is an investable asset, quoting him: "The reason for that is because it’s productive, it’s revenue generating, it is fungible, it’s used by just about every cloud service provider."
In the same CNBC-linked coverage, Emons is described as identifying China as the "single biggest threat" to the financing model, with the possibility that domestic production could trigger a price war that erodes collateral value.
Network bottlenecks and next steps
Data Center Knowledge warns that new AI campuses can create a separate requirement for physical routes that equipment upgrades cannot address, and it says new fiber routes can take more than a year to build and approach two years.
“new fiber routes can take more than a year to build and approach two years”
The same source describes a timing problem for AI developers, saying a data center can be complete, powered and equipped with GPUs while lacking the connectivity needed to operate as part of a larger distributed AI system.
Ron Westfall, vice president and practice lead for networking and infrastructure at HyperFrame Research, says existing routes are already running into capacity limits as AI factories move into new markets to access available power.
In parallel, Data Center Knowledge quotes Jimmy Yu of Dell’Oro Group saying he has not heard of wide-area network bottlenecks caused by AI traffic, while also noting that purpose-built routes with low latency and thousands of fiber pairs could become critical.
The financing plan’s stakes extend beyond chips and into infrastructure, as TechCrunch reports Nvidia’s guarantee is designed to address "wrong way" risk, where Nvidia’s obligations grow as demand weakens, potentially squeezing its revenues.