Elon Musk says Tesla will likely spend $3B to $4B on Nvidia hardware in 2024 and that half of its formerly noted $10B AI-related capex will be on internal spend
Context & Ripple Effects
Tesla’s disclosed AI budget shows that its compute buildout initially depended heavily on Nvidia hardware, while a substantial share of spending remained internal. That combination makes Tesla both a major accelerator customer and a company seeking control over the systems built around those chips.
Later coverage of Tesla’s long-term Samsung manufacturing agreement for AI6 chips illustrates the parallel path: buying leading compute where needed while developing more proprietary silicon capability. Nvidia’s expanded 2024 investment activity in core AI companies also places Tesla’s purchase within a wider effort to deepen the AI ecosystem that consumes its hardware.
First-order effects
- Tesla is likely to direct $3 billion to $4 billion of 2024 spending to Nvidia hardware, providing Nvidia with a sizable customer commitment tied to Tesla’s AI plans.
- Half of Tesla’s cited $10 billion AI-related capital expenditure is earmarked for internal spending, directing meaningful resources to in-house AI infrastructure and development rather than hardware procurement alone.
Second-order effects
- Tesla’s split between external accelerators and internal spend strengthens the case for a hybrid hardware strategy: Nvidia captures near-term demand, while Tesla retains incentives to reduce dependence through its own systems and chips.
- Large customer deployments make Nvidia’s hardware ecosystem harder for alternative accelerator vendors to displace in the near term, even as customers invest in proprietary designs.
Third-order effects
- If this pattern persists, AI compute buyers may increasingly separate immediate access to merchant accelerators from longer-term ownership of custom silicon and infrastructure—a direction Tesla later framed as potentially requiring its own chip fab.
- The resulting market could remain concentrated around leading accelerator suppliers for fast deployment, while major buyers gain more leverage and differentiation through vertically integrated hardware over longer cycles.
The trend: AI-intensive companies are pairing large near-term purchases of Nvidia compute with longer-term investments in proprietary infrastructure and silicon.