“Nvidia lines up $500 billion in financing as CEO Jensen Huang tells CNBC his chips are ‘investable asset’”
What actually happened
Nvidia signed memorandums of understanding with six major asset managers, Apollo, Blackstone, BlackRock, Brookfield, Goldman Sachs and KKR, to mobilize third-party capital for AI infrastructure buildouts. This matches Nvidia's own newsroom release confirming "strategic partnerships to establish independent compute financing platforms with Apollo, BlackRock, Blackstone, Brookfield, Goldman Sachs and KKR to mobilize over $500 billion of third-party capital for the buildout of AI infrastructure over time". Huang is pitching this as proof his chips are now a bankable, depreciation-resistant asset class comparable to real estate or toll roads.
Key facts
- Confirmed target: over $500 billion in third-party capital, per Nvidia's own press release, not just CNBC's characterization.
- Six named partners: Apollo, Blackstone, BlackRock, Brookfield, Goldman Sachs, KKR, all via MOUs, not signed, binding deals yet.
- The deal is described in the article as designed to let customers borrow against compute "much like commercial real estate, toll roads or other assets", an analogy from Nvidia/Huang, not an independent valuation.
- The article discloses the deal follows "a July swoon in global markets in which investors began asking whether Big Tech's AI investments would pay off", and that "rating agencies like Moody's have warned that unprecedented capital expenditures are beginning to squeeze free cash flow and force tech giants into heavier debt loads".
- BlackRock's Fink compared the scheme to "the creation of mortgage-backed securities in the 1970s", itself a loaded historical parallel given how MBS ended in 2008.
What to watch for
Watch whether these MOUs convert into binding financing agreements with actual terms, or stall at the announcement stage. Also watch for independent analysis of GPU resale/residual value as newer Nvidia chip generations ship, since that's the load-bearing assumption behind calling compute a "long-lived" asset, and whether any of this financing structure ends up funding purchases of Nvidia's own hardware (a circularity concern raised elsewhere in AI-infrastructure financing).
