Why everyone from OpenAI to SpaceX is building their own chips (and turning up the heat on Nvidia)

Summary

Nvidia has dominated the AI chip market for years, but the era of total dependence might be ending. OpenAI just shared its plans to spice things up with Jalapeño, its custom inference chip built with Broadcom, joining Google, Apple, and SpaceX in a growing list of companies building their way out of single-supplier risk. The goal is less of a reliance on Nvidia and more control over costs and supply chains.

Why it matters

The shift toward custom silicon signals a major transformation in the AI hardware landscape. For years, Nvidia's GPUs have been the de facto standard for training and running large language models. However, as AI scales, the cost and availability of Nvidia chips become critical bottlenecks. By designing their own chips, companies like OpenAI, Google, and SpaceX can optimize for their specific workloads, reduce per-inference costs, and secure their supply chains. This trend could erode Nvidia's market dominance and spur innovation in chip design, benefiting the entire AI ecosystem.

Related tools

• AI chip comparison tools
• Hardware benchmarking tools
• Supply chain analysis tools

Impact on AI tools/models

Custom chips will likely lead to more efficient and specialized AI models. For example, OpenAI's Jalapeño chip is designed for inference, which could reduce the cost of running models like GPT-4. This may enable broader deployment of AI in consumer and enterprise applications. Similarly, Google's TPUs have already powered models like Gemini, and Apple's chips optimize on-device AI. As more companies build custom hardware, we may see a fragmentation of the AI chip market, with different chips optimized for different tasks—training, inference, edge computing, etc. This could accelerate the development of smaller, more efficient models that run well on custom hardware.

What to watch

• AI chip market trends
• Nvidia's response to custom chip competition
• OpenAI's hardware roadmap
• Broadcom's role in AI chip design

FAQ

What is OpenAI's custom chip called? OpenAI's custom inference chip is called Jalapeño, built with Broadcom.

Why are companies building their own chips? To reduce single-supplier risk and achieve cost savings.

Which companies are mentioned as building custom chips? OpenAI, Google, Apple, and SpaceX.