NVIDIA and Broadcom shares face headwinds as OpenAI pivots to a multi-supplier strategy, prioritizing purpose-built inference chips to reduce long-term operational costs.
OpenAI's CFO confirmed supplier diversification, reducing NVIDIA's pricing power and future inference market share.
Broadcom dropped 7.95% as its custom-accelerator revenue is concentrated and sensitive to any single lab's ramp pause.
AMD held flat as its Instinct roadmap and deals with Anthropic and Meta validate its position, but not directly impacted by OpenAI's comments.
OpenAI CFO Sarah Friar confirmed a shift toward supplier diversification, signaling that the AI giant no longer views NVIDIA as its sole default provider. With $122 billion in cash and no immediate IPO pressure, OpenAI is leveraging its financial independence to optimize inference costs, putting downward pressure on NVIDIA's pricing power and Broadcom's custom-accelerator revenue.
By treating NVIDIA as one of several qualified suppliers rather than the default, OpenAI is compressing NVIDIA's pricing power and potentially limiting its share of the high-growth inference workload.