Wall Street's biggest private capital houses have thrown their weight behind NVIDIA in a $500bn move to make compute an asset class. The company signed memorandums of understanding with six firms to set up independent compute financing platforms, treating NVIDIA's chips and systems as investable assets to fund AI 'factories' for frontier labs, enterprises, and cloud providers.

With tech companies expected to spend over $730bn on AI this year, demand for computing power exceeds hyperscalers' funding capacity. Banks remain wary of large loans against assets lacking long track records, leaving private capital to fill the gap. "We began by building chips; today, we are helping create a new class of productive, investable infrastructure: AI factories," said NVIDIA founder and CEO Jensen Huang.

Apollo's Jim Zelter described modern compute as a scarce, mission-critical asset class poised for long-term economic growth. BlackRock's Global Infrastructure Partners arm, Microsoft, and MGX launched the AI Infrastructure Partnership in September 2024, chasing $30bn equity and up to $100bn with debt, recently leading a $40bn acquisition of Aligned Data Centers.

Brookfield plans a $100bn data centre campus in Kentucky, while KKR anchors Helix Digital Infrastructure, which NVIDIA seeded. NVIDIA has invested up to $100bn in OpenAI and holds stakes in Core Weave and Nebius, though Huang insists these represent only a small fraction of customer funding needs.